Modeling Permanent Magnet Effect on Resonant Cavity in HFSS

Hello,

I'm modeling a simple cylindrical resonant cavity with waveguide excitation of the TM010 mode. It takes about 30 seconds to run this simulation, no problem, plenty of experience doing this type of thing.

Now I want to see the effect of placing a permanent magnet at the bottom of the cylinder - I have a magnet from a microwave oven magnetron, found a source that says it has a strength of .19 Tesla.

So I created the donut shaped magnet at the bottom of the cavity (its about 60 mm OD, 24 mm ID, thickness of 13 mm), set it to be ferrite material with magnetic saturation of .19 T. Then I set this component to have an excitation>magnetic bias>uniform bias with 151,000 A/m.

Now the simulation doesn't solve - well it has been running for about an hour now. I tried setting the maximum number of mesh elements inside the magnet to be 10,000 - but no luck speeding anything up.

Do permanent magnets require THAT much more computation? It seems unreasonably long which makes me think I'm doing something wrong.

Thanks,

Kurt