Bow-tie active FSS, simulation convergence issue - modeling advice
Dear All,
I am currently studying a bow-tie FSS from an paper (two dimanond-shapped conductive surfaces). I have tried to model it, especially from the tutorials about the Jerusalem Cross FSS and the patch antenna on finite substrate. I draw the corresponding polylines in the susbtrate (which is very thin = 0.127 mm ; region : custom dieletric Rogers 5880), then making them PEC.
As it is an active surface, I added a load (1.8 Ohms) to a port between the two surfaces as illustrated in the screenshot (along with the model file). I managed to place the port in the middle on a wire, not touching the substrate, as issued in the forum (a post about feeding).
For the simulation parameters :
- PBC with repsect to the dimensions of the substrate
- frequency range of [3 GHz - 7 GHz] with interpolation (the surface should resonate at near 5 GHz)
- plane wave source
- standard meshing, 0.01 mm for wire radius
- CEM validation is OK.
Asking for : Transmission/Reflection coefficients.
Results : it takes a long time, and warns me about convergence issue ('WARNING 31096: ADAPTFEKO has reached minimum frequency stepping, results might not have converged').
Can you give me an advice to fix this problem ? If it is a bad modeling, or bad parameters used ? At least, can you validate this model ?
Thank you.
NB : I tried a version without the wire and the port, and whenever is the PBC spacing, it gives me grating lobes on the predicted resonning frequency range (Edit : even with PBC spacing > and >> lambda/2)
NB: I am new to FEKO (~ 2 months).
<?xml version="1.0" encoding="UTF-8"?>
Answers
-
Hi AurelienSurier
You can set the continuous frequency convergence to high.
The substrate is very thin. I would set a local mesh on the bow tie wings - 4x the substrate height is a good rule of thumb.
<?xml version="1.0" encoding="UTF-8"?>
0
