Learning Forum

[fparveen]

Hello,

I am simulating a system having an antenna in front of a head model. I am running simulations for 2.85 - 3.55 GHz at separate single frequency setups with frequencies at 0.1 GHz interval. The antenna dimensions are 30mm x 44.4 mm. So the far field at 3.5 GHz is at 4.6 cm. I have placed the head model at 4.5 cm away from the antenna. So that it is in the far-field. I have placed radiation boundary at lambda/4 distance from the antenna radiating planes (lambda at 2.75 GHz).

I am having the following issue: for frequencies above 3.25 GHz, the S11 for the system with head model is coming smaller than the S11 without the head model in the system. But with the head model the reflections should be higher. Can anyone help me figure our why this is happening? I am attaching the S11 plots for the system with head model and for antenna only.

(In the figure, 'ant' means antenna only simulation. '0 to 80' means I have rotated the antenna at different angular positions around the head, i.e., 0 degree, 10 degree etc., keeping 4.5 cm distance from the head, when there is a small blood clot placed inside the head. 'without blood' means when there is no blood clot inside the head. Here the S11 for 'ant' should be less than the other plots where there is head in front of the antenna. But it is showing small reflections when the head is removed form the system, keeping all the other setup the same, for example the radbox size.) What can be the reason for getting smaller reflection when there is head model, at frequencies above 3.25 GHz? The presence of head model should produce more reflections at every frequency.

[AndyJP]

>I have placed radiation boundary at lambda/4
It is a minimal acceptable distance for non-broadside. I would recommend larger odd number of ╬╗/4, and even larger at the broadside.

[Praneeth]

It looks like the setup is not proper. That could be the reason for improper results. Please check thoroughly.
You can also run a parametric to move the head in your model and analyze the results.
All the very best.

[fparveen]

Hello Thank you for letting me know about the radiation box size requirements. My antenna is a Vivaldi which has end-fire radiation. I have kept the radbox size lambda/4 at the non-endfire direction only. In the end-fire direction the radbox is much larger than lambda/4 because I need to accommodate the head there. However, in order to verify, I have increased the radbox size greater than 3*lambda/4 toward the non-endfire direction as well, and checked the S11 of the antenna. It gives almost similar results as the smaller radbox. It is shown below-
Since, they are close, seems my antenna is not producing much radiation in the non-endfire direction. So I am not discarding all my results with previous radbox. But, I still am confused about the question that I asked at the beginning of this thread, why at some frequencies, the S11 is greater in the antenna only case, when the head is not present?
Please suggest if I am correct to expect that dB(S11) should be greater at each frequency when the head is present, than in the empty radbox.
Or, is only the magnitude of the S11 not saying much about the amount of reflection it is getting, as there is the phase of S11 which also bears information?

Time domain:
To check, I performed inverse Fourier transform ifft(S11) , using the complex S11 so that both the magnitude and phase of S11 is considered. I am attaching the real part of ifft(S11) plots below.
I can see bigger amplitude for the case when the head is present than the empty radbox for 0 - 1.5 ns. Here the head is located at a distance of 4.5 cm from the antenna. So, first reflection from head should come at time T. Where T = antenna travel time + round-trip travel time for 4.5cm.
Here round-trip travel time for 4.5 cm is 2d/c = 2*4.5cm/3e8 = 0.3ns.
Can anyone suggest how should I calculate the antenna travel time? My antenna length is L = 30.2 mm and uses dielectric constant, er = 10.2 for substrate. So should I calculate as follows-
antenna travel time = 2*L*sqrt(er)/c = 0.643 ns ?
If it is correct, then T = 0.643 ns + 0.3 ns = 0.943 ns
So, I should get same amplitude until T = 0.943 ns , because the first head reflection should not arrive before that. Am I correct? If yes, then why I am getting larger amplitude before 0.943 ns for the case with head?

Thanks for reading the long post. I would highly appreciate any help because I am really stuck at this concept.