Elmer Discussion Forum

I have problems with my transient 2D induction motor simulations when I am using newer devel copies. Nonlinear system doesn't converge in MgDyn2D and crashes in next solver which try to use that solution. I have tested direct umfpack and iterative methods.

I tested few different commits and it seems that this problems started in Oct 11 commit 1212858e9b2915b3940e296a9d87efb788eb1b2f. Earlier Oct 7 commit 1212858e9b2915b3940e296a9d87efb788eb1b2f works just fine.

I am running virtual Ubuntu 22.04 inside win 11 wsl2.

Is there something that I should change in the solver settings when using these newer commits or is there something broken in the solver?

Hi, This was just fixed in "devel" branch. It was introduced when adding some superconductivity stuff to circuits. -Peter