Learning Forum

[jinan.nijem]

Hi 

I need help with the bandstructure simulations.. I tried to download two examples and do some changes to suit “ my structure” but I am not getting reasonable results.. I am trying to reproduce a published paper result. 

 Firstly, I tried to solve the problem using the Lumerical “mode” software. I followed up the following example:

https://optics.ansys.com/hc/en-us/articles/360042039033-Bandstructure-of-planar-photonic-crystal-with-a-square-lattice-varFDTD

 

My structure is a 1D photonic crystal which consists of a waveguide with one of its side walls looks like a sinusoidal wave. 

So, after importing the structure, I changed the boundary conditions to be pml  in the y-direction, while keeping Bloch boundary condition in the x direction and I give the normalized  kx=0.25

Based on https://optics.ansys.com/hc/en-us/articles/360041566614-Rectangular-Photonic-Crystal-Bandstructure

 

k _norm =k_SI * a /2/pi

 

k_SI=2pi/lambda, lambda=1550nm

a=400nm

 

 However whenever I run the program it gives me the window that the program is “updating slab modes” and it  stucks.why?

 

Second, I tried to solve the problem with FDTD and I followed the following example:

https://optics.ansys.com/hc/en-us/articles/360042038473-Bandstructure-of-photonic-crystal-waveguide-Line-defect with the same changes that I mentioned above

 And here’s what I got when I run the Gamma-X sweep from 0-1 

I tried to do many changes in the apodization width and center, to change the kx , changing frequencies ..etc. But all my trials doen't help.

So what are the possible changes that I can do to obtain a clear bandstrure? 

 

[Guilin Sun]

 “updating slab modes” is just a reminder message. No worry about it.

"Gamma-X sweep from 0-1 " it sould be from 0 to 0.5 since it is folded.

I would suggest you to continue using 3D FDTD as 2D varFDTD might not be that accurate.

I guess there are some unwanted frequency points on the "bands". Usually it means the simulation time needs to be a little bit long. But it cannot be too long to reach autsohutoff min default value of 1e-5. Otherwise the bands will have some missing points.

You may need to use symmetry-type of BCs to separate TE/TM modes if the structure has some symmetry: https://optics.ansys.com/hc/en-us/articles/360034382694-Symmetric-and-anti-symmetric-BCs-in-FDTD-and-MODE 

Other factors please read this post: 

Ansys Insight: Why my simulation result is different from published paper or experiment?