[zizou2020]

I'm not very familiar with FLUENT, typically just do structural modeling. I was interested in doing some simple models that involved some fluid capillary / meniscus effect.

I used this tutorial as guide. The approach taken in it is:

1) select multi-phase, and have two fluids in the material definition, air and a liquid (water in this case). Eulerian.

2) set surface tension values (0.072) and allow wall adhesion, then in boundary conditions add a contact angle (30) to the walls.

3) use adaption region->patch to make one part of the volume/area liquid, the remainder air.

So I attempted to begin this with basic versions to see if it works.

However, my results are quite inconsistent. I find that when I model a geometry similar to the one in the tutorial, or a geometry with two rectangles/areas, it appears to work as expected (or reasonably close).

However, when I have a structure with a single rectangle/area/whatever, it does not.

The settings appear to be the same on both (definition of the materials, gravity, mixture, etc). I'm probably missing something obvious. Help!

[ghorrocks]

Surface tension modelling is very sensitive to time step size. You probably just need to run with a smaller time step size. Have a look at the Courant number in your simulation and check it is not getting too big.

Also note surface tension modelling is super sensitive to mesh quality. Aspect ratios bigger than 1.5 or so will result in loos in accuracy of the Laplacian pressure. Try to use a 1:1 aspect ratio hex/quad grid.