I have written an indepth tutorial for the design of corrugated horn antennas and how they can be simulated with openEMS, and it was suggested that I show how symmetry planes can be used to reduce the size and time for the simulation.
To begin with I just wanted to add a PEC wall normal to the x axis located at x=0. I did this with the ve x boundary condition. I set the nf2ff not to include the ve x wall of the nf2ff box. I set the Calcnf2ff to include the mirror olane in the x axis.
However now the radiation patterns are not correct in one plane. They should look like this for 3 cuts.
But with one PEC symmetry plane I get this.
The script is this attachment. It's a modification to my original script without symmetry.
I would be very gratful if anyone can point out why im not getting the correct radiation patterns in all three cuts. I can't see what I'm doing wrong. Thank you.
Ps, in the tutorial I show equivalent simulation results from CST and HFSS and the results are almost identical!
Symmetry planes for horn antenna
Moderator: thorsten
Symmetry planes for horn antenna
 Attachments

 Corrugated_horn_with_mode_converter_options_Symmetry.m
 Modified script with symmetry
 (22.84 KiB) Downloaded 351 times
Re: Symmetry planes for horn antenna
Well I do not see an obvious problem. (except the missing mirror for the 3D far fields, but that does not change the cut planes)
I'm running it now, but it takes a while.
What you could try is to run the nonmirrored simulation, but with the cut nf2ff recording and apply the mirror then. This way should be clear that the fields and excitation and so on are fine. Even do this once for +x and x...
I'm running it now, but it takes a while.
What you could try is to run the nonmirrored simulation, but with the cut nf2ff recording and apply the mirror then. This way should be clear that the fields and excitation and so on are fine. Even do this once for +x and x...
Re: Symmetry planes for horn antenna
Hello Thorsten,
Thanks for looking into it. I'll try what you suggest. So if I understand correctly, I simulate the whole model as usual but then to implement the mirrror planes in the nf2ff function and check the patterns?
Thanks for looking into it. I'll try what you suggest. So if I understand correctly, I simulate the whole model as usual but then to implement the mirrror planes in the nf2ff function and check the patterns?
Re: Symmetry planes for horn antenna
I repeated the simulation with the full mesh and made sure that all planes for the nf2ff box are included (set to 1) with all boundaries set to PML_8. I left the CalcNF2FF symmetry plane on normal to the x axis with 'Mirror', {0, 'PEC', 0}. Just one symmetey plane for now.
This time the patterns look almost as the original, but I noticed some low level ripple on the phi=0 plot as shown in the picture.
Does this provide any clues as to what the problem might be?
Edit It also seems that the gain is 3dB greater.
This time the patterns look almost as the original, but I noticed some low level ripple on the phi=0 plot as shown in the picture.
Does this provide any clues as to what the problem might be?
Edit It also seems that the gain is 3dB greater.