Discussion Forum

I've got a question about how resonant cavity modelling is properly done in COMSOL (specifically 3D).

I'm modelling a cylindrical cavity and applying a high frequency signal by a small coax probe at the base. I've used a formula from the Balanis book "Advanced Engineering Electrodynamics" to construct a cavity of the correct size for the mode and frequency I'm interested in.

I've modelled the cavity first in 2D, then in 3D, using essentially the same method to model the coax cable as the coax-to-waveguide coupler example here: www.comsol.com/model/coaxial-to-waveguide-coupling-1863

Essentially, you hollow out the inner conductor so as to not solve for the field.

The 2D, axisymmetric model works quite well: at the resonant frequency, the field strength in the cavity is high and the reflection coefficient low(ish) (s11 =~ -10db)

The 3D model does not: at the frequency of the highest field strength inside the cavity, the reflection coefficient is still quite high (s11 =~-0.007 db). Note that the field strength is actually higher in the 3D case, but somehow more power is being reflected.

I have some additional modelling that requires a 3D model to work correctly, and am beyond confused as to how I am seeing the results that I do in the 3D model. I've attached both versions. I've omitted some of the lossy boundary conditions & materials in the 3D model as they shouldn't have an impact on the fundamental behavior of a resonant cavity. I don't think I've failed to translate any of the important elements...would appreciate any input.