No convergence achieved when refining internal isoparametric coordinates

[selrahc]

Hello,

I’m running a gem simulation based on the example found in Examples/Elmer/lem. I modified the gmsh code so I could get a double and a triple gem foil cell. When I simulate an avalanche with the single or double foil geometry I get some nice avalanches, but when I run the triple foil geometry and get the following warning message and the simulation doesn’t seem to move on:

ComponentElmer::Coordinates13:
No convergence achieved when refining internal isoparametric coordinates
at position (0.00213346, 0.00276998, -0.202893).

The position showing above is located inside the gas volume.

What does this warning mean and how can I fix it?

[hschindl]

Hi,

this message tells you that the calculation of the local coordinates within an element did not achieve the usual precision. This can be due to very much curved elements. In such cases, the first order approximation is used - this is usually good at the percent level.

Since this function is called many, many times it is harmless to have a few 100 of such messages. You can turn off the warning using ComponentFieldMap::EnableConvergenceWarnings(false);.

Having lots of those messages can point at poor meshing or (near) conflicts of volumes.

[selrahc]

Thank you. I waited about 6 hours and the program did show an avalanche. I guess I’ll have to go over the meshing.