[chegdan]

I'm attempting to adapt the interFoam solver to incorporate cyclic boundary conditions. I have had a look at the ChannelOodles solver and added several files (mainly the createGrad.H and writeGradP.H). The solver was able to compile and I have run the damBreak case with no problems, receiving similar result to the interFoam solver. However, when I apply cyclic BC's to another case, then there are problems. I must choose a ridiculously small courant number (1e-5 or less) to even have continuity converge and even with that the pressure gradient still diverges. I was unable to attach such large files (a mere 250 kb) so here is a link where they can be downloaded. Please take a look at it, see if there are any issues and maybe make some comments.

Thanks for any help that can be given. Also, I will be sure to share any knowledge gained with the community.

Dan

[chegdan]

Hello again,

I seem to be posting a lot lately, so I'll just continue to do so until I figure out my problem. Let me get to it. I'm trying to make a multiphase channel flow solver called rasInterChannelFoam (it borrows from rasInterFoam and channelOodles). So far I have been able to run cases where all of the fields agree with the cyclic boundary called "inOut" except for the gamma field. I have a simple case that as a block of water approaches the lower boundary, it is supposed to past through and fall from the top boundary. However, when the brick contacts the lower boundary, it compresses and then only passes a small fraction through the interface to the top boundary. Here are two pictures taken during the hit and just after the contact with the boundary.





I have been able to simulate single phase flow (with and without turbulence) that shows vortex shedding and field gradients (U, k, epsilon, etc.) pass through to through the cyclic boundary. However, as you can see there is no gamma passed through. Does anyone have a suggestion as to how I can correct this? Do I need to change the gamma equation to no include the compression term (just a thought although I'm not really sure how that will affect anything) or do I need to define another boundary type/condition to pass gamma? Lastly, is this a bug? I am using OF-1.5.


Procedure for generating mesh in Gambit and importing into OF:

1.Draw geometry (a box)
2.link top and bottom edges
3.mesh the object
4.define the top and bottom edge as one boundary called "inOut" of type "wall" and the sides as one boundary called "wall of type "wall".
5.Export the 2D mesh as .msh format.
6.Use fluentMeshToFoam utility to import the mesh
7.go into the polymesh folder and change the "inOut" to base type "cyclic".
8.After that I make sure that Ubar in the transport properties file has a low value that matches the initial uniform field of U.
9.Run my case.
10.Watch the crazy brick of liquid bounce back.

Here is a link to the video on youtube so you can see for yourself. Any Help is much appreciated. Here is a link to my zipped solver and test case.

Dan

[chegdan]

I finally found the solution to this one, but never replied again to my own thread.

http://www.cfd-online.com/Forums/ope...tml#post335817

and

http://www.cfd-online.com/Forums/ope...interfoam.html

better late than never i guess. I know there are other answers out there, but this is just closing this thread.