monopole antenna simulation

[xianyang gao]

Dear Everyone,

I recently simulated monopole antennas (shown in the picture) in both gprMax and CST. I found the o21 results of two monopole antennas facing each other in the air at a distance 10 cm are quite different in gprMax and CST, while the CST result matches the measurement much better.

I tried to think whether it is the material properties that caused such problem. However, when I remove all the other materials and only kept 'pec' triangle antenna and circle ground in both gprMax and CST, the trend remains the same - CST result is more similar to the experiment.

Any idea what is the problem in my code? or is there some problems for gprMax to simulate the monopole antenna (feed point problem)?

In both gprMax and CST, I put a 50 Ohms port between triangle antenna and circle ground.

Thank you very much.

Best regards

Xianyang

[Antonis Giannopoulos]

Hi Xianyang,

Very interesting comparison. CST has a number of advanced tools that gprMax does not implement (yet, or even ever!) Do you use a coaxial feeding source in CST? or is that a simple voltage gap source with 50 Ohms resistance? The retransmission line model that gprMax is not really a good choice for a monopole. It is easy to get a coaxial feed but is not yet available. Other points could be the conformal meshing option that CST has (i.e. reduce stair casing effects) and how thick are the "thin" materials. Are they simulated in each case with the same FDTD cells? Finally, we need to make sure that the distance to the PMLs are the same (i.e. the two models are similar in size) 

Best

Antonis

[Craig Warren]

Hi Xianyang,

A couple of things I noticed about your simulations that might be causing problems:

1. Have you checked the geometry of your model thoroughly? In the input file you attached, you specify the triangle command (which creates the monopole) before you place cylinders for the ground plane and plastic mounting. So from what I can see, you are overwriting (and thus destroying) the triangle, thus not simulating a monopole.
2. It also looks like the monopoles and feeds/receivers in the model are not in the same position as the real antenna - in the model they are a significant distance from the ground plane, yet in the real antenna it looks like there are only the thickness of the plastic from the ground plane. So you should check this also.
3. Can you plot the comparisons on a single graph. Even the measured data looks different on the gprMax and CST plots.

We have carried out (and published) comparisons for dipole antennas in free space and over half spaces, modelled using gprMax (FDTD), CST (FIT), and time domain integral equations (TDIE). The match between gprMax and CST was very good, you can see the results at - http://www.jpier.org/PIERM/pierm57/18.17021602.pdf

Also, as Antonis pointed out, you must ensure you are comparing apples with apples. So in your case, using the same feed mechanism in each model, and ensuring CST is not using any conformal meshing.

Let us know how you get on.

Kind regards,

Craig