Calculating transmission and reflection coefficients of a multilayered structure

Yes I have set everything. Please tell me how to set up TEM mode for waveguide port used for excitation.

I've try to simulate your problem in faster and not fine simulation and the results are:

S11=S22=-180dB
S12=S21=-95dB


When you start the time domain simulation check the bock "modes=All"

It seems to be Ok. But I can't understand why its coming wrong in my case. Can you sent it to me please?

Sorry but i'm not in office now. Try to sent my configuration:

- Brick dimensioni: W200; H300; t=3 (mm)
- Background: Normal; 0mm in all direction
- Material: Air
- Frequency Range: 0-2 Ghz
- Boundary: E=0; H=0; Open Add Space
- Port: two waveguide ports in opposite direction
- Start: mode selected= all
- Accuracy: -50dB

I think that's all. ok?

Marco

Pity you don't have FEKO - this type of model can be easily done with a reflection and transmission coefficients calculation using Periodic Boundary Conditions.
Below I have modelled 3 infinite dielectric layers stacked on top of each other with dielectric parameters similar to human tissue. The layers can have arbitrary thickness.

Hi Mona7150

Do you want to analyze and calculate the reflection and transmission of a plane wave which incident into a multilayer thin layer ? If, yes.

you can easily do it by TFCal software or other thin film softwares ... استخدم احدى هذه البرامج

zionico90 said:

Sorry but i'm not in office now. Try to sent my configuration:

- Brick dimensioni: W200; H300; t=3 (mm)
- Background: Normal; 0mm in all direction
- Material: Air
- Frequency Range: 0-2 Ghz
- Boundary: E=0; H=0; Open Add Space
- Port: two waveguide ports in opposite direction
- Start: mode selected= all
- Accuracy: -50dB

I think that's all. ok?

Marco

Click to expand...

Hi zionico90

Thanks. It does work in simple air slab. But when I add one more layer i.e some dielectric layer sandwiched between air slabs, the simulation takes a lot of time (hours together) and it also shows the warning that "current simulation will take very long time...." I tried changing mesh settings but nothing is working out. Also does the dimensions of the slabs have any effect on the results and simulation?

- - - Updated - - -

@sim4me

Thanks but I don't know how to use FEKO. Can you give me some tutorials so that I can design the problem in this?

@efendi

Thanks. I am not sure what thin film layers are. The thicknesses of the layers in my case are 10,20,90,3 mm. Can this software be used ?

Mona7150 said:

Hi zionico90

Thanks. It does work in simple air slab. But when I add one more layer i.e some dielectric layer sandwiched between air slabs, the simulation takes a lot of time (hours together) and it also shows the warning that "current simulation will take very long time...." I tried changing mesh settings but nothing is working out. Also does the dimensions of the slabs have any effect on the results and simulation?

- - - Updated - - -

@sim4me

Thanks but I don't know how to use FEKO. Can you give me some tutorials so that I can design the problem in this?

@efendi

Thanks. I am not sure what thin film layers are. The thicknesses of the layers in my case are 10,20,90,3 mm. Can this software be used ?

Click to expand...

Hi @Mona7150

please send your CST file for me , then i will try to simulate your structure.

I think the best method for you is the transfer matrix method (TMM). I have created good notes and even recorded lectures on this method, but I am not allowed to include a link to this on edaboard because they consider that "self promoting." Do a search for "EM Lab" at the University of Texas at El Paso. It is under "Resources" and then "Computational Electromagnetics."

@sme9572

PFA the cst file. This is only a one layered problem which I have started as an example.

rrumpf said:

I think the best method for you is the transfer matrix method (TMM). I have created good notes and even recorded lectures on this method, but I am not allowed to include a link to this on edaboard because they consider that "self promoting." Do a search for "EM Lab" at the University of Texas at El Paso. It is under "Resources" and then "Computational Electromagnetics."

Click to expand...

Thank you rrumpf. I will go these lectures.

I am new here, your topic caught my attention- I actually search for something similar to what i am doing, and yours is similar to mine.

I did modeled and designed some similar things to what you're describing in CST, I have no idea of IE3D. Straight to my thinking of what i feel you're asking of, i used CST 2013 to design the models- but it seems CST on support 3D designs. My dimension of the thorax of an human is still 100*100mm and thickness of each tissue of the thorax was specified in the z-direction, with values from articles. Dispersion values were then fixed by specifying the frequency of simulation.

To calculate my scattering parameters, S11 and S21, I put two antennas at opposite sides of the layered structures and created individual ports for the antennas. Afterwards, since i am interested in port 1(s11 and s21) data only, I simulated the transient solver by activating only port1 and calculating the scattering parameters for port1 only.

Present stage I am is to actually verify the accuracy of my simulation, because my first layer is skin and being a very lossy material, it seems all power from the antenna were dissipated at the skin layer.

Let me know if you have more understanding of what i did, we can get to share ideas

- - - Updated - - -

I have similar issues as @Mona7150, can you be of help? I can send my CST files and provide further explanations of what i intend to achieve. I have carried out the simulations myself, but i need a guarantee that what i did is okay. Still learning to use CST and EM simulation software generally.

Thanks in advance.

Hello oluwatosin,

I was also facing the problem of validation of my results in CST, so I shifted to MATLAB now.
Whatever I have tried must be clear through all my posts. I didn't tried anything with antennas,maybe it works well. Is your structure 1D or a 3D? If its 1D then there are other methods to deal with the problem.

Hi Mona7150, I went through the posts and if your material is isotropic, you can do it easily with finite, differential-eqs based solvers.

You can use periodic B.C. on X+X-Y+Y- directions and open on Z+Z- directions. This will mirror your slabs into infinite size (equivalent) and it will be a 1D problem. NOTE you should not use planewave, since planewave will be added to all the 6 walls of your domain and it will have issues. You need to use waveport, from that you can directly obtain your S11 and S21 and this is by theory equivalent to your infinite slab problem. All EM books talk about such slab problem at first class.

Alternatively, you can also solve this problem using integral-eqs based solvers, such as MoM. However, there is one problem, you have to get the Green's function of your system first before you can calculate the field. Freespace greens function is very eazy, But multilayer Green's function is not. You have to be careful there.

All in all, this is assuming your material is isotropic, and the wave propagation direction is along Z, perpendicular to your XY slabs. If you have an oblique incident angle, things will get complicated if you material is anisotropic.