Discussion Forum

I want to model a thin lossy dielectric layer within the RF module. The goal is to study the role of geometry of an RF cavity to maximize the quality factor given
1) a lossy conductive material (currently modeled using an impedance boundary condition)
2) a thin layer of dielectric on the surface of the conductive material with a given loss tangent

The rational is that the dielectric will have maximal loss where the local field is largest while the conductor will have maximal loss where the current is largest. I unsure of the best approach to modeling 2). I though I might be able to modify the material definition for the impedance boundary condition by introducing a complex relative epsilon and mu but this dint seem to have any effect. I'm assuming this is because the impedance boundary condition represents a thin metal surface with no internal E and B fields.

In any case, I'm now thinking of introducing a thin 3D layer inset inside the cavity geometry and assigning a lossy dielectric to that domain. However it seem like a pain to generate this domain and depending on how thin I make it I expect it to add considerable computational overhead due to small mesh features required to treat it.

I'm wondering if I can define a 2d shell instead formed by e.g. just making a scaled copy of the current cavity geometry and assigning a dielectric loss to the shell. Will this work? If I could do that I'm not sure how to define the surface model appropriately either. would I use an impedance boundary condition? and if so should I expect a complex epsilon or mu to correspond to a lossy dielectric?

I've attached the cavity resonators example from the application library as a reference starting point.