[Eterson]

Hello everyone,

I’m working on a VOF droplet simulation with `interFoam`/`interDyMFoam` and using `dynamicRefineFvMesh` to adapt the mesh around the interface. My goal is to obtain a **thin refined ring around the droplet interface** while keeping the interior of the droplet and the exterior gas **coarse (unrefined)**.

Here’s a simplified version of my `dynamicMeshDict` when using gradient mode:

FoamFile
{
version 2.0;
format ascii;
class dictionary;
location "constant";
object dynamicMeshDict;
}

dynamicFvMesh dynamicRefineFvMesh;

dynamicRefineFvMeshCoeffs
{
field alpha.water;
mode gradient;

refineInterval 1;

lowerRefineLevel 1000; // refine if |grad(alpha)| > 1000
upperRefineLevel 90000;

nBufferLayers 1;
maxRefinement 2;
maxCells 3000000;

correctFluxes
(
(phi none)
(alpha.water none)
);

dumpLevel true;
}


The issue is:

* If I use very low thresholds (`lowerRefineLevel ~ 1e-12`), I get refinement — but only on the *gas-side offset* of the droplet contour, not symmetrically around the interface.
* If I set more realistic thresholds (e.g. `1000`, based on `magGrad(alpha.water)` values), then the log reports:
* Selected 0 cells for refinement out of X
* Selected 0 split points out of a possible 0 and no refinement happens at all.

but when I see the values on paraview of grad(alpha.water) it shows a field much bigger that 1000 , as you can see in the image. I even check the range of grad(alpha.water) with a fixed apropieate refinement and show the a similar range.


Do you guys know what I'm doing wrong? Any advice, or example configurations where gradient-based refinement successfully unrefines the interior, would be greatly appreciated.

Thank you very much,