I'm coupling the electrical equation (superconductor in ac fields), which generate heat, to the heat transfer equation. The electrical equation uses curl, linear, elements, due to the special properties of the conductor. The thermal equation uses second order discretization, as default. I found I can never get the model converged. But after changing the thermal equation to use linear elements. The model converged.
I'm confused and wonder whether coupling physics using different orders of elements has some problems in principle? Or is this trivial? Anyway I hope I can use second order elements for the thermal equation in the future, which can give better resolution. But I can't get it converged yet.
I'm confused and wonder whether coupling physics using different orders of elements has some problems in principle? Or is this trivial? Anyway I hope I can use second order elements for the thermal equation in the future, which can give better resolution. But I can't get it converged yet.