Simulating point source in HFSS

Hi people,

https://www.edaboard.com/viewtopic.php?t=95356

In this topic above jackle asks 2 questions. But as i see the first question is left unanswered.

As jackle asks in his first question, I wonder if it is possible to simulate an "ideal" point source in HFSS. To be more specific, let me introduce my problem:

I have an infinite ground plane and on top of that a very thin dielectric slab. What my supervisor asks me to do is place a point source at the [air/slab] boundary (also at the center of the slab from the top view) and calculate the fields starting from approximately 5λ away from this source. (all this trouble is to calculate the surface porpagation constant BTW ) He also told me that if it is not possible to simulate a point source in HFSS (he does not know much about the program indeed) i could use a very small (infinitesimal) dipole.

As to re-ask jackle's first part of the first question, is there an efficient way of simulating a point source in such a case or should i just use a dipole?

For the second choice what do you suggest me to do? Place a dipole made up of two cylinders or two rectangles? If I use cylinders should i embed half of the cylinders inside the slab? How should I give the excitation? Use a rectange between dipole arms and create a wave port or lumped port? Or feed this rectangle with a probe placed just at the bottom of the inf. gnd plane? Or... is it possible to use a single rectangle and give a current excitation along it? (i.e. a monopole antenna).

i don't know very much about the ports and excitations. if you can enlighted me i will be very happy and grateful.

Thanks in advance.

IMHO, in order to have the excitation at the interface, you should draw two rectangular sheets of small width and length about λ/50 at the interface and parallel to it with small separation and connect the two by a lumped port (gap voltage source). Hope this works. Else you could search on how dielectric surface waveguides are excited and use a suitable excitation.

-svarun

1. You can't get ideal point source results from HFSS.
2. if you are interested only at 5λ away, a lumped source applied on a small area is more than good enough.
3. You would have a hard time extracting surface wave propagation constant from the field distribution. because, A) 5λ is too big to handle comfortably with HFSS; B) even if you get the field, it would not be only the surface wave
4. For 1 layer case, it is much more convenient to solve the dispersion equation to find the constant directly..

Thank you all, four your ideas and suggestions. I will take them into account. Later I will share the results and maybe ask a few more questions. Thanks.:D