[miyazawa]

I want to refine my unstructured grids (hexahedron only) according to gradient of p, but I am not sure whether I have to obtain the gradient files firstly. I use topoSet, refineMesh, and mapFields to fulfill refinement, and successfully refine a bounding box. I feel confused that the gradient files form in post-processing like the walGradU, or can be achieved by modifying several parameters.
Please help me to refine my unstructured grids.

[cutter]

Hi,

please have a look at the tutorial case .../tutorials/multiphase/interDyMFoam/ras/damBreakWithObstacle (checked it for 2.4 and later versions). In this case the mesh is refined based on the phase field alpha.water at each time step (refinement defined in constant/dynamicMeshDict). I haven't tried this yet, but I'm pretty sure this can be done based on the pressure field p with only minor modifications.

Hope that helps.

Cutter

[emjay]

I know this thread is old, but for other user who also searching for a workaround.

tutorial case .../tutorials/multiphase/interDyMFoam/ras/damBreakWithObstacle
would not help for this problem. The dynamicMeshDict file refine a scalar only between two values. It is not possible to define a gradient field. e.g. grad(p).

It is interesting why there isn't any workaround here to define a meshRefinement on a gradient basis.

Or maybe i haven't figure it out who to do it.

[Wikie]

Mesh refining based on a gradient is a topic, that I'm also interested in. If anybody has a solution for this problem, please let us know

[clapointe]

A quick work around is taking the magnitude of the gradient field to produce a scalar field. You could also use a single component of the field.

Caelan

[gaza]

Originally Posted by emjay

I know this thread is old, but for other user who also searching for a workaround.

tutorial case .../tutorials/multiphase/interDyMFoam/ras/damBreakWithObstacle
would not help for this problem. The dynamicMeshDict file refine a scalar only between two values. It is not possible to define a gradient field. e.g. grad(p).

It is interesting why there isn't any workaround here to define a meshRefinement on a gradient basis.

Or maybe i haven't figure it out who to do it.

Hi
here it is shown how to adapt solver and case to refine mesh based on a gradient:
http://www.tfd.chalmers.se/~hani/kur...rayDyMFoam.pdf