This example is inspired by the conclusion of this thread on the Elmer forums.

The geometry comprises:

• a 3x1x1 cuboid of 'steel', modelled as a material with relative permeability of 5000
• a closed 'coil' around (and conformally touching) the steel
• an airbox filling the rest of space

The original issue in the thread was finding that the Fix Input Current Densities on the WhitneyAV solver would create non-zero current density in the air and steel. This resulted in unphysical "vortices" in the calculated B-field within the steel.

The important aspects of resolving this case were:

• Use the Fix Input Current Densities feature of the CoilSolver, and disable it on WhitneyAV
• Fillet the coil geometry to avoid sharp corners; otherwise the solver would not converge

Calculated coilcurrent e:

Calculated magnetic flux density e: