CFD Online Discussion Forums

CFD Online Discussion Forums (http://www.cfd-online.com/Forums/)
-   OpenFOAM Running, Solving & CFD (http://www.cfd-online.com/Forums/openfoam-solving/)
-   -   twoPhaseEulerFoam - Help needed (http://www.cfd-online.com/Forums/openfoam-solving/108821-twophaseeulerfoam-help-needed.html)

Mineral November 2, 2012 04:20

twoPhaseEulerFoam - Help needed
 
Good Morning/Evening to you Foamers

I am currently trying to set up a case using the twoPhaseEulerFoam, and I am running into all kinds of trouble. Due to lack of documentation I seek your help here!

The case is a simple flow through a duct with a little wall-thingy in the middle, with air as the continous fluid and water as the dissparsed.

The geometry was made in a CAD software and meshed with Netgen.

I made a simple inlet and outlet BC's
Inlet zeroGradient pressure, outlet is fixed 0 pressure.

Inlet 1 m/s airflow in x direction, outlet zeroGradient.

It is probably the easiest way to explain is if you can have a look at the case folder attached.

https://www.dropbox.com/s/zsugcgly6u...aseCase.tar.gz

Now the problem is that the water phase is just stuck in the inlet, what do you think might be causing this?

Some additional questions:
Some the of the parameters of the solver are also a mistery to me and I would like to ask if any of you can explain the following:

1. In trasportProperties what does the d mean? Accoring to the bubbleFoam wiki, it is a phase diameter, but what does that mean in practice? Some sort of artificial particle diameter?

2. In the same file alphaalpha, what does that represent?

Does anyone know where I can fint a paper on the model so I could read up on the effects of parameters such as the virtual mass and ect?

3. Does anyone have a paper on the k-epsilon turbulance model? The parameters implemented in my case folder are just some values taken from the tutorials, and actually that might be the problem with my simulation altough my money is on the BC's.

4. Last and final question (for now) is that how can I make sure that the scale of the model is correct? That is that openFoam/Netgen treats the geometry in the same dimensions as the CAD software which is in mm?

Thanks for your attention

BT

sharonyue November 2, 2012 22:47

Quote:

Originally Posted by Mineral (Post 389810)

1. In trasportProperties what does the d mean? Accoring to the bubbleFoam wiki, it is a phase diameter, but what does that mean in practice? Some sort of artificial particle diameter?

as far as I know, the da and db is the diameters of the two phases. if phasea is a air bubble , the da is the average bubble diameter. if the phaseb is water, think about this: if the air was injected high flow rate.the water would create a water droplet, so the db is the diameter of the droplet.


Quote:

Originally Posted by Mineral (Post 389810)
3. Does anyone have a paper on the k-epsilon turbulance model? The parameters implemented in my case folder are just some values taken from the tutorials, and actually that might be the problem with my simulation altough my money is on the BC's.

in the tutorial, u can find the turbulance is off, Prof.alberto said the turbulance model in this solver is not perfect.
http://www.cfd-online.com/Forums/ope...eulerfoam.html

thats what I konw, correct me if i was wrong.

Mineral November 7, 2012 09:37

Thank you Forrest for your input, I have now found the solution to my problem and it was the scaling of the model. I checked the points file and the domain was from (0 0 0) to (50 50 100) but using transformPoints -scaling '(0.001 0.001 0.001)' did the trick!

I also found some handy info about k-epsilon here on cfd-online http://www.cfd-online.com/Wiki/Standard_k-epsilon_model

To bad though that the model is not good in twoPhaseEulerFoam

Also I found somwhere that alphaalpha is not used in the solver.

I have a question now tough, has anyone knowledge or experience how to implement a BC where a boundary acts as an outlet for phase A, and a wall for phase B? In someway like a filter?

Or some other option to make part of the boundary absorb the dispersed phase.

sharonyue November 7, 2012 20:11

Thank you for that informtion.

type: inletOulet would be suited for you. more details you could check it out in the user guide.

Best,

Mineral November 11, 2012 21:40

I have now tried many different methods and lots of google searches, but still no success. inletOutlet does not do the trick although it sounds logical to me.

sharonyue November 11, 2012 21:50

Quote:

Originally Posted by Mineral (Post 391537)
I have now tried many different methods and lots of google searches, but still no success. inletOutlet does not do the trick although it sounds logical to me.

can u upload your case?

Mineral November 12, 2012 02:36

1 Attachment(s)
Quote:

Originally Posted by sharonyue (Post 391538)
can u upload your case?

Yes, please have a look at this, this is a simple 2D test case with coarse mesh.
The inflow is air-water mixture with 1% water, flowing through a channel with two wall where their front faces are supposed to absorb the water. The BCs in the attached case are just simple zero velocity on all wall faces, but I have been playing around with the BC on the two faces which I want to absorb the water, called "filter" in the boundary list.

Cheers, hope you will have time to have a look at this.


All times are GMT -4. The time now is 21:32.