Note: This discussion is about an older version of the COMSOL Multiphysics® software. The information provided may be out of date.

Discussion Closed This discussion was created more than 6 months ago and has been closed. To start a new discussion with a link back to this one, click here.

3D bandgap analysisof photonic crystal: eigenvalues are wrong

Please login with a confirmed email address before reporting spam

Hi,

I try to work from the 2D model (bandgap analysis of photonic crystal in model library) to the 3D case for a long time .

My model is a simple cubic with a cyclinder inside. I have set floquet periodic boundary conditions in all direction with kx ky kz . I also copy the meshes for the periodic boundaries pairs to avoid divergence due to boundary meshes.

Now the problem is that all the eigenvalues I get for Lamda is zeros(extremely small) for both the real and imagary part. (lamda should be very big for the imagary part!) . If I choose a reference eigenfrequency to calculate the eigenfrequency as the 2D model, no matter which frequency I choose, all the eigenfrequency i get is nearly the same the reference eigenfrequency(thus no gaps in all regime).

Did anyone could kindly give me some suggestions or simply share a successful 3D bandgap analysis example for photonic crystal? thanks a lot in advance! my email is lcs1354@hotmail.com

cs

2 Replies Last Post 10 avr. 2014, 04:29 UTC−4

Please login with a confirmed email address before reporting spam

Posted: 1 decade ago 22 nov. 2010, 07:18 UTC−5
Hello cs,

I'm also trying to move from the example of the 2D case to the 3D but I'm completely stuck. I use the matlab script provided that I'm slightly adapting to making it work in 3D but it seems it doesn't work.

Have you been able to fix your problem? From what I have read around on the web there is two points to pay attention to. First, as you did, to copy the mesh among periodic boundary pairs (It doesn't work in my case, I got an error when I try to do it. How did you manage to make this working??). Second, the eigenvalue linearization point that helps greatly the solver. It should be different from zero, better if big, and best if complex. Last, it seems it is better to have a good eigenfrequency intial value, in the worst case try to sweep it a bit manually by 2 or 3 orders of magnitude if you don't know what you are looking for exactly.

Also how did you define your global equation? In 2D it is freq=1-nEz because TE and TM are decoupled but what about 3D? I tried freq=1-nE with nE=(Ex*conj(Ex)+...)/A but not convinced at all. Boundary conditions are like you.

Could you or anyone else around that has any idea on how to solve some of the issues I am encountering give me some advice on this 3D bandgap analysis problem.

Many thanks,

Yan (l.pech@hotmail.fr)
Hello cs, I'm also trying to move from the example of the 2D case to the 3D but I'm completely stuck. I use the matlab script provided that I'm slightly adapting to making it work in 3D but it seems it doesn't work. Have you been able to fix your problem? From what I have read around on the web there is two points to pay attention to. First, as you did, to copy the mesh among periodic boundary pairs (It doesn't work in my case, I got an error when I try to do it. How did you manage to make this working??). Second, the eigenvalue linearization point that helps greatly the solver. It should be different from zero, better if big, and best if complex. Last, it seems it is better to have a good eigenfrequency intial value, in the worst case try to sweep it a bit manually by 2 or 3 orders of magnitude if you don't know what you are looking for exactly. Also how did you define your global equation? In 2D it is freq=1-nEz because TE and TM are decoupled but what about 3D? I tried freq=1-nE with nE=(Ex*conj(Ex)+...)/A but not convinced at all. Boundary conditions are like you. Could you or anyone else around that has any idea on how to solve some of the issues I am encountering give me some advice on this 3D bandgap analysis problem. Many thanks, Yan (l.pech@hotmail.fr)

Please login with a confirmed email address before reporting spam

Posted: 1 decade ago 10 avr. 2014, 04:29 UTC−4
Even for 2D PhC I have such problem... After searching around no useful solution is found. Anyone has an idea about this problem?

Thanks!
--
Pu, ZHANG
DTU Fotonik
Even for 2D PhC I have such problem... After searching around no useful solution is found. Anyone has an idea about this problem? Thanks! -- Pu, ZHANG DTU Fotonik

Note that while COMSOL employees may participate in the discussion forum, COMSOL® software users who are on-subscription should submit their questions via the Support Center for a more comprehensive response from the Technical Support team.