[ingabobjoe]

Hello, I'm working on a corrosion model using a DEFINE_GRID_MOTION UDF in Fluent to control the inward movement of nodes and faces in a metal sample based on interaction with a flowing fluid. I'm primarily using the remeshing dynamic mesh method, which allows me to specify a minimum and maximum cell size to tell the software when to remesh. The remeshing is going well until a certain time when one of the cells kind of collapses instead of remeshes (see images), and I receive a "negative cell volume" error. The smallest side of the collapsing cell (example shown in the image, although the cell location changes depending on my settings) is clearly below the cell size I specified since other cells are getting remeshed, but I'm wondering if the cell size is being measured with the maximum rather than the minimum length.

Most solutions to this error I've read online have encouraged me to decrease the time step, but I've tried and have encountered the same issue regardless, I think because the complexity of my mesh motion makes such thin cells problematic in certain cases. It wouldn't be a problem though if my mesh would remesh those cells according to the parameters I gave it. Does anyone have any insight into this issue?

pre-error.png

post-error1.png

post-error2.png

mg_remesh_settings.PNG

mesh_movement.jpg

[ingabobjoe]

Please let me know if further information would be helpful!

[ingabobjoe]

I should also mention that a frequent issue is the development of really tiny cells (well below the minimum length scale) that aren't getting remeshed even though they should be as far as I can tell, since other larger cells are being remeshed elsewhere in the domain. I've attached a couple images here (particularly the second) that illustrate this, where the second image is a zoomed in look at the first. These cells typically cause "Negative cell volume detected" errors during the following iteration.

mesh_overview.png

mesh_smallcellissue.png