CFD Online Discussion Forums

CFD Online Discussion Forums (
-   OpenFOAM Running, Solving & CFD (
-   -   cyclicACMI boundary condition treatment (

jimteb March 10, 2015 13:08

cyclicACMI boundary condition treatment
Hi All,

I was wondering if anyone could enlighten me on how the cyclicACMI boundary condition introduced in OpenFOAM 2.3.x works, more specifically:

-How are two boundary conditions handled on the same cell face? e.g. when there is partial overlap of a cell face (with say a wall BC) with another mesh region where fluxes are being interpolated to/from?

The reason I am interested in this is I am trying to use this BC for a compressible flow model with rhoPimpleDyMFoam. I am consistently getting oscillating pressure fluctuations at the region connecting the two non-conforming meshes, maybe suggesting some sort of conservation problem ...

Any comments would be much appreciated,


wyldckat April 24, 2015 16:15

Greetings James,

I know I wrote about this the other day here on the forum... ah, found it:

Originally Posted by wyldckat (Post 540179)
I did some checking of how it works and wrote it here: - essentially there are actually 4 patches needed for each "cyclicACMI" pair or patches: 1 wall and 1 cyclic for each side.

Mmm... this doesn't answer your question. If I remember correctly, the problem with the current ACMI implementation is that it assumes/enforces that the wall part of the pairs must act as a convention wall boundary:
  • no-slip condition (flow speed always set to 0);
  • with zero gradient for pressure, as well as for all other fields.
This is why there are a few bug reports about not being able to define other boundary conditions for the wall part of the patch pairs.

What this effectively/probably results in is the situation where the ongoing fluid flow hits a wall when it wasn't expected, i.e. there is an unexpected backward step or forward step that occurs when the faces are only partially covered. Oh, and it should be using AMI interpolation for weighting the overlapped faces, hence using both wall and cyclicAMI for weighting the exchange between the two sides.

Beyond this, I'll have a look into your other post on this topic, since it might be easier to diagnose the problem that way:

Best regards,

All times are GMT -4. The time now is 23:10.