THz FSS (frequency selective surface) simulation with CST
时间：03-30
整理：3721RD
点击：
Hi,
I'm trying to simulate a frequency selective surfaces in CST MWS.
I am using Frequency solver with floquet mode boundaries, meaning I choose unit cell for X,Y and open (add space) for Z.
Right now I'm only interested in transmission coeff. for zero incident angle, so I select only the first 2 floquet modes TE(0,0) and TM(0,0) for excitation (corresponding to a plane wave).
Now, my problem is the following.
When I simulate from 1 to 20 THz I get a different result than going from 1 to 10, then from 10 to 20 and then just putting the second graph at the end of the first.
I added a picture to show what I mean.
Attachment 53046
The top pictures are the single 1-10 and 10-20 measurements. The middle one is the combination of them.
The bottom picture shows the full 1-20 measurement, missing the two additional peaks at 8.5 and 9.5 THz.
Could anyone try to explain why this happens. And why only for 1 to 10 THz, because the 10-20 part looks pretty much the same in both calculations.
The results, in general, also seems to differ from the literature (see second attached picture).
Attachment 53047
Peak transmission frequency at 100% is the same, but where do the sidelobes (90% !) in CST come from? Due to the above problem I cannot really trust the CST calculations. Who tells me I chose the correct interval?!?!
By the way, the structure (for testing purpose) is a simple 33x33um square aperture in an infinitely thin 51x51um copper sheet. So not at all some kind of a "complicated" structure, but like the simplest case possible.
I hope anyone can help me. I'm almost sure I'm only doing a very basic mistake, assuming CST does not deliver incorrect results for such simple calculations ;)
I'm trying to simulate a frequency selective surfaces in CST MWS.
I am using Frequency solver with floquet mode boundaries, meaning I choose unit cell for X,Y and open (add space) for Z.
Right now I'm only interested in transmission coeff. for zero incident angle, so I select only the first 2 floquet modes TE(0,0) and TM(0,0) for excitation (corresponding to a plane wave).
Now, my problem is the following.
When I simulate from 1 to 20 THz I get a different result than going from 1 to 10, then from 10 to 20 and then just putting the second graph at the end of the first.
I added a picture to show what I mean.
Attachment 53046
The top pictures are the single 1-10 and 10-20 measurements. The middle one is the combination of them.
The bottom picture shows the full 1-20 measurement, missing the two additional peaks at 8.5 and 9.5 THz.
Could anyone try to explain why this happens. And why only for 1 to 10 THz, because the 10-20 part looks pretty much the same in both calculations.
The results, in general, also seems to differ from the literature (see second attached picture).
Attachment 53047
Peak transmission frequency at 100% is the same, but where do the sidelobes (90% !) in CST come from? Due to the above problem I cannot really trust the CST calculations. Who tells me I chose the correct interval?!?!
By the way, the structure (for testing purpose) is a simple 33x33um square aperture in an infinitely thin 51x51um copper sheet. So not at all some kind of a "complicated" structure, but like the simplest case possible.
I hope anyone can help me. I'm almost sure I'm only doing a very basic mistake, assuming CST does not deliver incorrect results for such simple calculations ;)
I'm wondering if this could have anything to do with the mesh model being automatically adjusted for different frequency regimes?
i cant open the attachment. can u reattach. I think its because of the mesh. Meshing will be done based on highest wavelength. So the mesh will change for 1-10, 10-20 and 1-20 GHz. Check the "Mesh Properties" option. There you can see a option "steps per wavelength".