[2538sukham]

Greetings.

I am simulating a piston cylinder assembly using foamExtend3.2 by modifying topochangerFvMesh class linearValveLayersFvMesh removing the sliding and attach/detach. On a 2D geometry which is pure Hex mesh created by blockMesh, the interface capturing is ok as seen in figure topoChanging2D.

solver : interDyMFoam

The problem arises in 3D model where the interface is wiggling resulting in unphysical interface as shown in the next figure. This unphysical behavior occurs when the interface compression is set to 1 and the error amplifies with increase in the cAlpha. Changing the fvSchemes is of no help. I believe this has to do with the VoF on topoChanging meshes with layer addition/removal.

My question:
1. Is it better to use the dynamicFvMesh solver such as displacementLaplacian to move the patch i.e. the piston and ten map it to a new mesh and continue? This will increase the simulation time. If so, how to proceed with this? (Like the movingCone simulation in OpenFOAM v11 meshToMesh) .

2. The simulation works fine until the time reaches a certain value after which the interface starts wiggling. How do I quantify the error involved? (e.g. plotting the pressure before and after topoChange maybe?)

checkMesh report is attached.
The linearPistonLayersFvMesh.C used for the topology change is attached.


With regards!