How to set radiation boundary in HFSS?

Hi!

I'm newbie with Ansoft HFSS and trying to simulate lambda/4 microstrip antenna. Until now I have

- drawn a plane on outer cross section of microstrip feed for waveguide port (not assigned to any material. should it be?)
- defined the plane to waveguide port and defined terminal line from the top of the groundplane to the bottom of the microstrip (center of the strip)
- drawn a sphere (radius ~lamdba/2) which completely encloses the whole geometry (at least ~lambda/4 away from model) and defined radiation boundary to it.

When I'm trying to run simulation, HFSS claims:
port refinement process abc3d: Port 'Waveport1' is incorrectly set up. A port must have none existent material on one side. Both "Sphere1" and "Sphere1" exist.

Can I just draw a box to which I define radiation boundary? Can the waveguide port be located on the radiation boundary?

OK, I guess I have already managed to do some things. It seems that you can't consider waveguide port in calculation domain but lumped port. Also it seems that waveguide port cannot be on same plane as radiation boundary. Maybe it's also better to use rectangular box as boundary rather than sphere?

Any further advices?

Thanks,
Markus

Hi,

does anyone know how fatal error these are:
[warning] Warning - Boundary 'Rad1', Boundary 'PerfE1' : Boundaries overlap. (3:10 heinä 06, 2004)
[warning] Warning - Boundary 'PerfE1', Boundary 'LumpPort1' : Boundaries overlap. (3:10 heinä 06, 2004)

Actually these warnings are from one of the HFSS example project which came in installation CD, so I guess this is not serious. But does anyone have more specific information?

Thanks,
Markus

Hi.
you can use followin terms:
1)Drawing Radiation box.(greater 3 or up order than Dimensions of your antenna.)
2)Giving any copy of other objects of your ant.
3)Subtracting copeis of objects from radiation box.
now creating bounderies.
1') select bounary
2') select radiation
3') click on sides your radiation box.

Don't worry about the Boundary overlap. Just make sure the more critical one is defined later. So it is kinda a priority sequence. The one defined later dominates.

Hi,

Firstly, Radiation boundary is meaning E-field going into the plane that you selected as Rad boundary perpendicularly. Lumped port is internal source port that the signal will radiate continuously along the port to your simulation structure. in which if when HFSS calculate the field strength from port to radiation boundary that is zero, that will not be the real case although it is okay for simulation. By the way, lumped port is supposed not to be in the bottom of radiation boundary, since it is internal source port, I think that you could use wave port in the bottom of boundary then u wont find any error.

Enjoy,
Le

Hi Markus,
1. The warning that you get that the boundaries overlap can be ignored. For example, if I'm simulating a CPW, it becomes imperative that the waveport and the radiation boundary overlap.
2. Yes, you can use a rectangular box for the radiation boundary condition and as I saw, you already know that it should be at least lambda/4 from the model.

Ananth.