question on Green's functions

[Demonis]

hamid_iran

I have to analyse a printed antenna (a dipole with a complex form on a finite thickness substrate without a ground plane). I have read a lot of IEEE AP articles about this thema but I can not find good Green's function for this case.

Thanks in advance for the help

 

[loucy]

ie3d green function for layered dielectric

it is unlikely to find your desired Green's functions directly from any paper. The best way to get the result is to use commercial electromagnetic field solvers such as Zeland IE3D, EMSS Feko etc.

 

[Demonis]

With a commercial software it is no problems. But for my thesis I have to make a mathematical model of such kind of antenna.

 

[khaira]

You can use Image theory to represent your Microstrip layout in free space. Then you can use Free Space Green's function.

 

[loucy]

The classical image theory is not applicable.

You can use the boundary element method, so as to avoid computing the Green's function for layered media.

The fastest way is to find some public domain code (mixed Finite element-Boundary elment) that can handle this type of problem.

If you have more than 1 year to finish the thesis, then you might start to "re-program" the Green's functions.

 

[Demonis]

Thanks for the help. I have enough time for "re-programing" the Green's functions (besides this is the main idea of my thesis). But I cannot find a good start point (a good example or book) for such kind of analysis. Can you recommend me something.

 

[loucy]

in that case, I would suggest the paper "Multilayered media Green's functions in integral equation formulations" by K. A. Michalski and J.R. Mosig, on IEEE-AP

 

[hamid_iran]

Derive the Green's functions yourself!
Just consider a current elemet on the surface where your printed antenna will resied, write the general form of the fields in substrate layer and upper and lower half spaces. And, finally applay the appropiate boundary conditions on the interfaces as well as radiation condition on infinity. Based on your method of formulation, electric, magnetic, or potential of this current element would be our Green's function.

 

[Demonis]

Thanks all for the answers. I am a newbie in electromagnetic analysis of antennas and will be very much appreciated for some more good examles (It can be something primitive (dipole antenna), but from the beginning )

 

[loucy]

In general, one should talk mainly with his/her advisor regarding the thesis topic. It is better to build on some previous work of your advisor and former students, especially because re-programming the Green's functions is time-consuming and not very "productive".

 

[Demonis]

The problem is that my advisor has experience only in analysis of waveguides and with antenna's analysis I should start from the very beginning, which is not so easy.

 

[loucy]

Regardless of the area of specialty for your advisor, you would choose a these topic that is new in a certain aspect. Maybe you have some new idea to analyze the antenna. But you must consider the fact that implementing even a simple idea require a lot of work--such as re-programming the Green's function. It might not be worthwhile.

The paper mentioned above is by far the clearest and most comprehensive on the Green's functions for stratified media. Take a look at it and see whether you like this stuff. Keep in mind that this is only a fraction of the information you need to implement a MoM code for analysis of Antenna.

 

[hamid_iran]

Dear Demonis,
As I have already mentioned, you can derive the Green's functions yourself!
The Green's Function derivation is not dependent on the radiating object but on the structure consisting of ground planes and dielectric layers. Therefore, the Green's Function for dipole antenna or any kind of colplex microstrip antennas are the same, provided that they have the same structure of ground planes and dielectric layers.
I am attaching an small file of my own work on deriving the Spectral-Dmain Green's functions for a strip line geometry, which consists of two ground planes and two dielectric layers (as you can see on the last page).
I hope it would help you to get started.

 

[Demonis]

hamid_iran Can you explain me why on page 14 you have a discontinuity of only Hy and not a Hx.

 

[gary977]

Please view the document "Modelling of Dielectric Materials in FEKO".

The techniques for modelling dielectric material in FEKO are currently: the surface equivalence principle (SEP), the volume equivalence principle (VEP), special Green’s functions, thin dielectric sheet approximation, dielectric coating for wires
and coated metallic surfaces/dielectric solids for Physical Optics. Each technique will be discussed below, before an example is presented.

 

[hamid_iran]

Dear Demonis,
Well, it is due to the B.C.'s on the interface.
As you know, anχ[H1-H2]=Js on the interface,
where, Js is the surface current on x-direction, an is the unit vector normal to the interface in z-direction, and χ means cross product. Therefore, the discontinuity would be on the y-component of the H, and x-component would be continuous.