plane wave in rsoft
Using CST-MWS , say , how would you generate the frequency response of a PBG showing band gaps. Would you use plane wave excitation or would you use waveguide ports . I need to see the response for TM and TE modes separately . How should I go about ?
-Arun
it depends on the frequency range of interest. If you are working in the microwave region, I would suggest that you use horn antennas in the simulation to get the transmission spectrum. The agreement with experimental results is great when you create the experimental environment in the MWS model.
That will take huge time I suppose. Do you know any other software which can produce gap maps using plane wave expansion technique ? Thanks for your time. There is this software called "Translight" posted on this forum a few weeks back. Have you used it ?
-Arun
Actually it doesnt take too much time. Except transmission through defect modes, the simulations take at most 1 day on a pentium-4 3 GHz.
You can use the MIT s photonic band gap package to generate band diagrams or Rsoft's Bandsolve. But the results does not essentially give you the transmission characteristics. Because there would be in general some bands which can not be excited by an outside excitation due to symmetry reasons.
PBG structure studying example of CST MWS.
Enjoy it.
Hi,
for a quick check you can use 2D calculations (with effective index approximation, if appropriate) and TE/TM excitations; you can use FDTD or FEM for this; i built a couple of FEM models in Femlab, if anybody is interested i will upload some here; in 3D with waveguide coupling structures, things are more complicated: you could use the waveguide mode(s) as input, but problem becomes quite demanding in resources and computation time (i managed only with FDTD here, FEM ran out of memory)
what kind of problem are you tackling, svarun?
regards,
baluba
I understand the PBG from EM point view. But in lots of literature, the figure plotted for PBG structure is always Frequency vs. L X M L. Can you help to explain it more details? Or can recommand some reference book.
Best Regards,
Those plots are actually k-vector versus eigenvalue plots. If you know solid state physics those plots will make sense to you. Since the photonic crystal is a periodic medium the wave equation for H-field turns out to be an eigenvalue problem smiliar to the eigenvalue problem of the electron in a real solid. Those X-L-M corresponds to the high symmetry points in the Brillouin zone. You can rather envision them as propagation direction inside the photonic crystal. Anyway, "Optical Properties of Photonic Crystals", by K. Sakoda is a very nice and elegant book on photonic crystals.
irfan1, thanks for you explanation and recomendation.
When analyzing periodic structure(photonic crystal), we always specify TE or TM polarization. I got confused about the TE/TM polarization here.
In electromagnetics, TE means no electric component in propagation direction, TM means no magnetic component in propagation direction.
But in PC studying(2-D infinitely periodic structure), TE incident means the E component paralle to the direction of rod, it is not consistent with the definition of TE wave in EM, am I correct?
You are right. The definitions are not the same that we use in waveguide theory. Even the TE/TM convetions used by different authors are not the same. TE/TM is used to tell the direction of electric field in the case of photonic crystals.
yes, by using this definition (TE polarization), what is the exact mode of the propapgation EM wave(in definition of electromagnetics)?
Best Regards,
Beware that in CST Microwave Studio, any photonic or electromagnetic band gap structures arranged in triangular unit-cell lattice will be model as an equivalent square lattice. In other words the natural symmetry points that the triangular lattice has will not be considered. This problem may be resolved by using HFSS and Rsoftdesign's BandSolve packages which an oblique unit-cell may be model.
Can you elaborate a little bit more on why this happens. So , do you think that if I model a microstrip patch antenna on such a substrate , I may not get the additional gain that is associated with the better band gap of a triangular lattice ? Please do clarify. Thanks for your time.
-Arun
TE, usually, used when the electric field is parallel to the axes of rods. But thats only just a convention. I use TM for the same thing in my papers.
For the case of the microstrip antenna, you will get the gain you want
On the other hand you can not obtain (simply) the bandstructure of a triangular lattice by using a rectangular unit cell. Because you will loose the rotational symmetry of the triangular lattice. But still it is possible.
irfan1, thanks for clarifying to svarun. Yes, I agree with you thus far. I also experienced a mixture of TE and TM conventions, especially in the text by Joannopoulos et. al. called Photonic Crystals and the tutorials in Rsoftdesign BandSolve.
And yes, modelling an EBG/PBG antenna consist of triangular lattices of array are no problem for CST solver as Unit-cell doesn't really come in to play. The surfaces waves simply travel in almost all directions, and it is up to the periodic structures rather they are of square or triangular arrangements to suppress those unwanted waves. Best regards!
By the way, deadline for submission in the IEEE MTT-S 2005 is December 1st. Hope everyone get some publications there.
Hi,
I recommend you to search for PBG in the upper right search box on Ansoft's website.
Presentation - Investigation of the Photonic Band Gap accelerator cells using Ansoft HFSS [PDF]
Presentation - Frequency Selective Surfaces [PDF]
Presentation - HFSS Vector Field Calculations [PDF]
Presentation - Design Applications of Defected Ground Structures [PDF]
Itai
the transliht software is helpful to you
https://www.edaboard.com/viewtopic.p...ght=translight
I found this discussion is really useful for me
I hope someone can lend me the CST example
urgent!
i dont have enough points
is that applied to FSS materials ?
Hi svarun
I think that in mws 2006 the triangular grid problem. has been solved
that's what the rep. said
PL