HFSS eigenmode solver for open structures

[Decesicum]

Hello,
can the HFSS eigenmode solver be used for structures which are not completely surrounded by metallic walls? (for instance a parallel plate waveguide, without side walls). If so, what type of boundary should I choose? I tried for rectangular waveguide and the resonance found are those of a rectangular cavity, not of the waveguide (I have TE101 mode, not TE10 mode). I tried to use PML but the results were irrelevant/not realistic.
Thanks,
D.

 

[nomigoraya]

Eigen mode solver is not the right choice to study waveguide modes you want as Eigen mode solver takes every structure as cavity and calculates the resonant modes as this is from where this name Eigen comes from. I think for your purpose simulate your waveguide with frequency domain solver. I think that will solve your issue

 

[Decesicum]

Thank you for your answer!
But how should I set the 'solution type" in this case? Because Modal/Terminal solution types do not allow to specify number of modes to compute and anyway I should add the excitation source... I use the rectangular waveguide just for testing purposes (if HFSS would return correctly its resonance frequencies, then I can rely on the solution it would provide for a real structure that I have and where I dont have formulas for the resonant frequencies, as I have for the waveguide...)
D.

 

[nomigoraya]

Since I work more with cavities and at the same time use CST more often so I am afraid that I may not answer it correctly for HFSS but I think in case of study for waveguides you need to specify both the entry and exit points to be ports otherwise no matter which solver you will use it will take it to be enclosed with PEC making it cavity again.
Once you define them as ports then I am sure you can tell the solver that in how many modes you are interested.
If I get sometime I will try to explore this with HFSS . One last point what is your problem I think you need not to be get confused with this all stuff

 

[Decesicum]

I want to see how much a car wheel does behave as a waveguide, where are its resonances.

 

[PlanarMetamaterials]

I attended the webinar release of HFSS 18 a few weeks ago, and specifically asked about this during their live Q and A.

18.1 is supposed to have what they call their "Characteristic Modes Solver", which is pretty much what you described -- an eigenmode simulator, but for open structures. Unfortunately, since 18 just came out, 18.1 could be a few months.

In the case of your PPW, PMCs on the sides (i.e., perfect H boundaries, not PMLs) should be fine, as that's how the structure is typically analyzed. The car wheel.... unfortunately, I can't really help you there.

 

[Decesicum]

Thank you for your answers! At least now I get correctly the cutoff frequencies for a rectangular waveguide (using Driven Modal, not Eigenmode). If it wont work correctly for more complex geometries, we'll wait till we get 18.1.
Thanks!
D.