Nothing touches the case. And they are pretty close hold on how the analysis works...
All my test cases look about like that. I imagine (not a physicist, so take it FWIW) the induction field has different density levels, and as the brass gets just to the melting spot, there are some hoter sections as it flashes just to the melting point.
Does not seem to be an issue when just going to annealing temps. The annealing appears to be nice and consistent.
ZY