HFSS Waveports vs Incident Plane Wave

[PlanarMetamaterials]

Hello,

I have recently been working with an incident pane wave excitation to obtain a TF/SF response of a simple scatterer, however, the data didn't appear to be correct, so I integrated the real pointing vector over a couple of orthogonal planes and used reflected1/incident1 and total2/incident1 to obtain the scattering parameters S11 and S21, respectively. These results gave me an entirely different response from the scattering parameters given by using waveports on the same orthogonal planes.

Has anyone come across this behavior before or have an explanation for this data discrepancy? Has anyone obtained scattering parameters with an incident plain wave before in a similar manner and had it work as expected?

Thanks.

 

[rodgerwxh]

If you use plane wave, I am assuming your problem is unbounded. Then how can you obtain S11, S21 for an open problem?
If you use the WP, what is your boundary condition surrounding the scatterer?

You did not describe your problem clearly.
NOTE: only parallel plate wave guide with two PECs and two PMCs can mimic a plane wave incident case. All other configurations plane wave will have modes to excite the structure, which is equivalent to WP.

 

[PlanarMetamaterials]

I suppose that could use some clarification.

The metallic scatterer is inside a PPW. In the case with the plane wave, the ports are replaced with rad boundaries.

 

[rodgerwxh]

I am assuming that parallel plate waveguide is setting top and bottom PEC and side walls PMC, along propagation directions is the rad BC. If this is the case, WP with first mode should be equivalent to plane wave incidence. You might want to verify this by checking the field at a location inside the PPW near the scatterer with both plane wave and port excitations, if the field is different, then I am doubting the setup is wrong. If the field are the same then I feel its more like a S11 calculation issue. Hope this helps.

 

[PlanarMetamaterials]

My setup is correct. It's the two different results that I'm having trouble with.

If the field are the same then I feel its more like a S11 calculation issue

Is there something wrong with the process I described in my original post?

The relative SF/TF field values are not the same as the waveport scattering parameters would suggest, hence my wondering if there is a difference that I am unaware of.

 

[PlanarMetamaterials]

For those who are interested, I found the solution to my problem.

In short, the problem was the second (receiving) port.

I had assumed that the second port (when not used as the source) would absorb all incoming fields, similar to a radiation boundary. It turns out that this is only true for fields which match the modal profile of the port setup - in this case, the port was expecting to receive a uniform TEM wave, but due to the scatterer, the impinging field was not uniform TEM, and so most of the energy was simply reflecting off of the port.

In the TF/SF simulation, there are no ports, so I was using a radiation boundary - which did absorb all incoming fields, as I expected. (Also of interest, but not related - I've learned that radiation boundaries only work in a vacuum - they also reflect when they are placed along a dielectric interface).

This was the source of the large result discrepancy between the two setups.