Proper initialisation of phase boundaries for interDyMFoam
 

Hello Dear Foamers,

I am facing currently with a problem, with proper setting of initial conditions for interDyMFoam solver. For my simulation I want/have to use dynamically refined mesh.

But the question is, if I can refine the mesh for time 0, to more accurately initialise the boundaries between two phases. The base coarse mesh is built from cubic cells. The shape which I try so set at the initial time with setFields is a sphere. Of course "the sphere" set by setFields in the relatively coarse base mesh is not a sphere, which influences the results of calculation substantially.

I've been trying such tools like refineMesh for mesh refinement in the region where the initial sphere is located, but interDyMFoam was refining the mesh at the interface again, which resulted in very small cells, which effectively stopped the simulation due to very low timesteps.

Is there any method of creating the dynamic mesh at initial time which could be used than by the solver?

Best regards,

Krzysztof

Hi,

I may be wrong, but if you used snappyHexMesh in parallel to generate your mesh (keeping both the outside and inside of the sphere, check the igloo tutorial for that purpose), your refinement level would be accounted for and you wouldn't have this re-refinement problem.

Best,

Pablo

Thanks for the tip,

currently I was working with cubic meshes due to ease of using it. I suspected that I missed something, and there is some tool in OpenFoam smarter than normal setFields which additionally will make this refinement.

Your idea of using snappyHexMesh for that purpose is promising. I will check it, thank you.

Regards,

Krzysztof

Hi,



How did you actually solve this problem in the end ? Using snappyHexMesh or which tool ?



thanks



Federico