Simulating a quarter wavelength monopole above an infinite GP

[Aspire-Design-Ltd]

Hi,
I am trying to simulate the standard quarter wavelength monopole above an infinite ground plane but I seem to be missing something.

The centre frequency is 2.4GHz but the impedance and radiation pattern are not correct.

I have a quarter wavelength wire above a PEC boundary. A lumped port has been added between the wire and the PEC boundary.

Do I need to add a metal ground plane or can a connection to the PEC boundary be used?
Thanks, any suggestions warmly welcomed!

MPC

FURTHER FINDINGS 1:

Monopole.zip

I tried extending the ground plane to 3 wavelengths.
I also changed the AddLumpedPort to AddCurvePort because I am using a 1D port in the Z direction.
I checked the paraview of the E and H fields and there is activity in the boundary PEC so this suggests you can connect to the boundary PEC as a ground plane.

The radiation pattern and impedance still do not make sense.


Monopole2.zip

FURTHER FINDING 2:
As suggested by @VolkerMuehlhaus I extended the port to two mesh lines. I also found that the AddWire function extends the wire beyond the coordinates specified.
Thus using a radius of 3 would short out the excitation port because the wire is extended. See the ParaView screen shots below. you can see a 31unit 3unit radius wire is 36units long.. The second screen shot shows the monopole extending into the PEC boundary layer.

By changing AddWire to AddCurve a single wire is correctly positioned.

From ParaView you now get a promising H-Field and RotH-Field. I am not sure what the difference between the two plots is:

The S11 and Impedance now look promising. As @VolkerMuehlhaus mentioned the far field radiation pattern does not look correct.

Monopole3.zip

Anybody any ideas on the problem with the radiation pattern.... Thanks.

FURTHER FINDINGS 3:
I tried extending the ground plane out from 3 wavelengths out to 5 and 10 wavelengths and from the screen shots below you can see the radiation pattern elevation is reduced as the ground plane size increases. Not sure if this is correct. Anybody able to comment?
Thanks.

[VolkerMuehlhaus]

I have a quarter wavelength wire above a PEC boundary. A lumped port has been added between the wire and the PEC boundary.

That's valid, but you should add more mesh lines at the port. openEMS wants to add a current probe in the middle between both port terminals, but you have only one (!!!!) cell across the port.

From what I remember, the NF2FF needs tweaking when some boundary is actually a reflector, like your PEC boundary.

I also changed the AddLumpedPort to AddCurvePort because I am using a 1D port in the Z direction.

Using the lumped port is fine.

---

If you are looking into wire antenna simulation, you might also like the yagi model that was discussed here:
#158 (comment)
That is based on openEMS using the Python interface, not Matlab/Octave.

[Aspire-Design-Ltd]

Hi @VolkerMuehlhaus, thankyou for the note about the port needing to add a current probe.
I also found that AddWire function extends the wire ends by an amount which appears to be equal to the radius. This causes serious problems for the unwary as the wire becomes longer than the defining coordinates....

I have updated the above with my findings and now have the S11 and feed point impedance looking reasonable but the Far Field is all wrong. Do you have any more links to the NF2FF issues with PEC boundary?

Greatly appreciate your comments thanks... I will keep on digging.....