Discussion Forum

Hi,

I'm trying to find the pressure inside a large, flat drop when it is evaporating. This involves solving

together with a decay condition at infinity. This much is fine, modelled by Dirichlet on a disk, and Dirichlet on a "large" sphere for the far-field condition, and agrees well with the analytic solution. The problem is, I next need to solve

in the disk, subject to Neumann constraints at the boundary. This is rife with problems:

• is zero exactly in the disk. I actually want uflux.u, but this is a post-processing variable, so apparently I can't use it in a calculation?
• I can instead locate a second "drop" a distance normal to the first, and calculate in there, but then Poisson doesn't converge using "Coefficient Form Boundary PDE" (no unique solution to Poisson with Neumann conditions - there is an additive constant)
• I've tried speficying a pointwise constraint , but then the solver just fails to satisfy Poisson close to that point
• I've tried using a Global constraint (), but I cannot seem to get this to work
• I've also tried a very small Dirichlet constraint close to , but then Poisson also seems to fail here!

Any ideas? This all has analytic solutions to compare against.