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/)
-   -   channelFoam (http://www.cfd-online.com/Forums/openfoam-solving/113816-channelfoam.html)

pedroxramos February 27, 2013 07:08

channelFoam
 
Hello there!

I'm doing a simulation of the flow around a cylinder with the solver channelFoam.

something like this: http://openfoamwiki.net/images/1/17/...rf_cylMesh.jpg

Any of you is working with a similar case?

I'm trying to adptate the channel395 example of openfoam but i'm not having success.

Any of you have a tutorial for thys kind of problem? Please, help me.

Regards.

wyldckat February 28, 2013 16:31

Greetings Pedro,

Can you be a bit more specific of what exactly you're trying to do?

Best regards,
Bruno

pedroxramos February 28, 2013 16:44

Hello Bruno.

I'm workin in a research project at FEUP, Porto.

We are trying to do a numerical simulation of the flow around a cylinder. Like this: http://www.youtube.com/watch?v=1u-Bx13PRGg

The numerical investigation will be validated agains experimental data of the open channel flow around a pille.

What kind of openfoam solver do you suggest? I'm trying the channelFOAM!

Regards.

wyldckat March 2, 2013 06:28

Hi Pedro,

The solver seems to be a good choice, but it depends on a few more details:
  • What have you already been able to do with OpenFOAM? If you're going directly into this solver, than it's a very big jump. It uses LES and it's one of the slowest kinds of simulations that can be performed for incompressible flows!
  • You can find a simple geometry for a flow around a cylinder in the OpenFOAM Programmers Guide, located in the folder "$WM_PROJECT_DIR/doc/Guides-a4". And more specifically, the case should be available in the "tutorials" sub-folder "basic/potentialFoam/cylinder".
  • You can setup more complex geometries if you use SwiftBlock. Or even use SwiftSnap if you have the channel geometry designed in 3D CAD.
  • Are you going to simulate only the submerged part?
Best regards,
Bruno

pedroxramos March 2, 2013 16:17

Hello Bruno!

I changed my plans. I think I should use the interFoam solver, using the VOF method. I have no relevant experience with OpenFOAM :(

I already build the 3D mesh with blockMesh.

I need to use LES model because I want compare the results obtained by Fluent.

Do you think interFoam is a good idea?

Thank you so much, Bruno.

Best regards.

wyldckat March 2, 2013 20:16

Hi Pedro,

At first I didn't even think that VOF supported LES... but then I remembered that the tutorials folder does have 3 kinds: "laminar", "ras" and "les".

I don't know which specific version of OpenFOAM you're using, but if you've got 2.1.1 installed, have a look into this tutorial that 2.1.x has got: https://github.com/OpenFOAM/OpenFOAM...s/waterChannel
You can then try and compare the case setup with the tutorial "multiphase/interFoam/les/nozzleFlow2D".

I honestly am not sure of what else to suggest to you. The rule of thumb with OpenFOAM is to gradually work your way up from a simple case to the target case. So try to keep the geometry as simple as possible to start with.
Starting with icoFoam, simpleFoam and/or pimpleFoam would help you get the feeling for how OpenFOAM works...
But if you want to stick to interFoam, then I suggest that you do some simple cases similar to your target case, starting with laminar, then RAS and finally LES, always comparing with Fluent if possible, so that you can get a good feeling on how things work on OpenFOAM. This of course is if and only if you feel more at ease with Fluent ;)

In addition, do not switch to parallel mode until you really need it! OpenFOAM's VOF in parallel requires some additional calibration of the "fvSchemes" file. Searching the forum for this information should bring up some possible solutions.

Best regards,
Bruno

pedroxramos April 11, 2013 06:01

Hello again!

I'm using the interFoam solver and I got tihis:
https://dl.dropboxusercontent.com/u/...ier9abril3.avi

I have a problem in the outlet. What boundary condition do you suggest?

Best regards.

wyldckat April 11, 2013 09:51

Hi Pedro,

Are you talking about the detail where the water level seems to reduce on the outlet only?

And what boundary conditions are you currently using? Are they identical (on the outlet) to the ones provided in the tutorial "waterChannel"?

Have you confirmed if the gravity "constant/g" is pointing in the right way?

By the way, I don't know if you have seen the following website: http://folk.ntnu.no/nilsol/cases/openfoam/ - it's geometrically different from what you're doing, but the type of simulation is similar....

Best regards,
Bruno

wyldckat April 20, 2013 14:06

Hi Pedro,

I know that you've been able to go further on this topic, at least from what I've seen from this thread: http://www.cfd-online.com/Forums/ope...tml#post421781

But I'm posting here about a very recent thread I've seen just now that might be useful to you: http://www.cfd-online.com/Forums/ope...nnel-flow.html

Good luck!
Bruno

pedroxramos April 23, 2013 05:33

Thanks Bruno!

Right now, the problem is the inlet zone.

As you can see here ( http://d.pr/i/yhog ) the water goes up in front of the inlet...

Regards.

pedroxramos April 26, 2013 05:54

Hello Bruno.

Do you know why the no slip condition causes this problem?


LIke it has been said here: I built a artificial inlet block with full slip condition in the walls and I got this


any suggestion?

mgdenno April 26, 2013 13:19

Hi Pedro,

Are you initializing the velocity field with setFields? I have seen that kind of wave when the inlet BC does not match the internal field.

Matt

wyldckat April 26, 2013 19:55

Greetings to all!

There is another possibility: the initial pressure fields might be incorrectly defined. Check the following thread for more information: http://www.cfd-online.com/Forums/ope...ity-field.html - more specifically, check post #7 ;)

Best regards,
Bruno

wyldckat April 29, 2013 17:51

Quick note: I found today the boundary condition:
Code:

variableHeightFlowRateInletVelocityFvPatchVectorField
I saw it on the list for "Inlet Boundary Conditions": http://foam.sourceforge.net/docs/cpp/a09425.html
Perhaps that boundary condition can help in not injecting too much water into the domain? I haven't re-checked the boundary conditions discussed on the related threads, to see if this is the one already being used...

JR22 May 8, 2013 06:27

Hi Pedro,

There was recent post that might interest you. It looks like a similar problem as the one you were having:

http://www.cfd-online.com/Forums/ope...t-k-omega.html


All times are GMT -4. The time now is 00:19.