[tecmul]

Hi all,

I'm running a moving mesh case with pimpleFoam and the SBRStress mesh motion solver in OpenFOAM v2006 that involves plate oscillation. At a certain point in the simulation when the plate nears its maximum angular displacement, the mesh starts to vibrate, as shown in the linked video. https://imgur.com/t8qj6XN
The solution blows up at the same time if the number of PISO correctors are below 3, and at least 3 non-orthogonal correctors are required as well. The attached image shows how the instability in the velocity magnitude mirrors that of the mesh motion.

Testing with different discretization schemes shows that, at a maximum courant number of 0.2, the only thing that prevents the problem occurring (or at least reduces the intensity of the vibrations), is a Laplacian scheme of "Gauss linear skewCorrected" for the cell displacement. But the problem persists if the maximum courant number is brought below 0.2, and the cells aren't even skewed in my simulation. I'm relatively sure that the problem is in the discretization of the cell displacement, as the location and extent of the vibration is altered by a change in that and nothing else. Even the most stable numerical schemes still result in divergence if the number of PISO and non-orthogonal correctors are below a certain limit.

I don't understand why this is happening. Does the SBRStress solver become unstable at some point? If so, why? Checkmesh reports that the maximum non-orthogonality of the mesh is around 68 degrees, which while not good, isn't terrible either. I'm hoping someone has seen something like this before and can offer some tips.

unstable solution.jpg