Method of analyze SIW antenna

[Nano_o]

dear all,
I want to know which method is more adequate to analyse the characteristics of a SIW antenna and what are its advantages compared with the rest of the methods?
many thanks!

 

[rrumpf]

That is a long story and we need to know more about your antenna. If you want a concise overview of a bunch of different methods, take a look at Lecture 1 here:

**broken link removed**

Attach a pic of your antenna, materials, etc., I think we could be of more help.

 

[Terminator3]

SIW (Substrate Integrated Waveguides) is one of my favourite topics. Theoretical approach is the same as slotted waveguide antennas. Formulas are the same. The only different place is transition. Microstrip line to SIW transition is separate research topic, but there are many online materials freely available. So for calculations all waveguide theory formulas can be used (SIW as DFW dielectric filled waveguide).

Software approach: Fastest way is to use 3D finite element solver.
You may be interested in this:
http://www.comsol.eu/model/substrate-integrated-waveguide-leaky-wave-antenna-16021
http://www.comsol.eu/model/download/281071/models.rf.siw_leaky_wave_antenna.pdf
(pdf can be downloaded freely from the first link). Tutorial looks nice, unfortunately no lite/demo versions. So i tried only FEKO LITE (out of memory, maybe enough to see single slot at low frequency, around 2..5GHz) , SONNET LITE (pretty interesting! I got S11 for slot that looks very realistic. Single SIW slot at 5GHz, memory is enough, but no radiation pattern in lite version).
I was very suprised that SONNET LITE can analyze SIW structure. For example:
Tunable SIW Bandpass Filters with PIN Diodes (using SONNET!)
**broken link removed**
Although it is not SIW antenna, i am not sure if there far field would be nice. But my S11 result for single slot seems in good agreement with theory (slot size changes S11 position, center frequency).

 

[Nano_o]

I didn't build an antenna yet but it will be a slotted one (simple and not stratified).
Now, I am confused about the adequate method that could be applied to get the characteristics (phase constant, attenuation,..etc), as going through the literature most used ones are MoM and Mode Matching, transverse resonance method, FEM but unable to decide which one needs to be chosen and implemented.
@rrumpf: thank you for the link unfortunately the codes are protected !
@Terminator3: thank you for the links, you didn't mention CST and HFSS any faced issues with those two? hope you could help with your experience in the SIW field.

Kind regards!

 

[Terminator3]

I did not used cst and hfss. What do you mean by phase constant and attenuation? To excite slots inphase DFW formulas can be used to find wavelength in DFW (dielectric filled waveguide). Then series fed slots each wavelength. Power division between slots in array made using the same approach as for slotted waveguide antenna arrays: using slot offset from symmetry axis, or slot under some angle to waveguide axis. You can start with formulas for DFW. There are more than ten papers on this topic with some formulas, googled and freely available online. I mean siw antennas. I usually add this text to search string: filetypedf , it helps to search only for pdfs

 

[rrumpf]

I have a list of free and open source electromagnetic simulation software about 2/3 down here:

**broken link removed**

For SIW, I think finite-difference and finite element methods will serve you best.

I have one more suggestion, consider studying your device in 2D first. The simulations will be much faster and you can get your device 90% designed very quickly. Move to 3D only as the last step.

 

[Nano_o]

Thank you, very much for the open sources, I appreciate it!
could you give any arguments for your choice or references , I will be grateful!
because I have been looking into the literature for a quite long period and I can't decide maybe because I am new to this field !
regards!

 

[rrumpf]

I understand your frustration. First, any method can be made to do anything so somebody experienced enough at a method could argue why it may be the best. Second, I see finite-difference and finite element methods as the purest solutions to Maxwell’s equations. It seems like most other methods take advantage of some sort of approximation or simplification. For example, to simulate metallic structures, the method of moments is probably superior. To simulate long periodic structures, semi-analytical methods may be superior. To simulate all-dielectric structures, Fourier methods tend to be superior. Time domain methods tend to be better for transient and nonlinear devices. Frequency-domain tends to be better for single frequency simulations or highly resonant devices. You can get most of these details and more in Lecture 1 here:

**broken link removed**

I am not completely sure what your SIW device looks like, but I do not any simplifications to the problem. That is why I pointed you to finite-difference or finite element. I like finite-difference for simulating things in new ways because I can formulate and implement new codes very quickly. Finite-differences also seem to be better right now for simulating very large structures and parallelization, but I don’t think this is a problem for you. Finite element can make more effective use of unstructured grids so will be more efficient for you in the end.

 

[Nano_o]

many thanks for the link.
Okay I will go for it, and go back to this post if any problems are faced
regards!

 

[soundav]

As per my opinion, FEM method is the best option for SIW antennas.
Now there is student version for FEKO which is far more better than the previous LITE version.
Its easily accessible too.
You can have better options than the LITE version

 

[anilkrpandey]

Attaching one design of SIW Patch antenna that is solved using FEM solver ( EMPro)



Design- View attachment SIW_Patch.pdf

Project- View attachment SIW_patch_empro.zip