Learning Forum

[minusunny]

Hi ,

I am trying to do Band structure analysis using  https://www.lumerical.com/learn/video/swg-bandstructures/ this example. When i try to calculate the effective index of first order TE mode (neff1) it is giving me a wiered result. My fundamental TE mode value looks okay.

my grating period is 222 nm and duty cycle is 50%. My waveguide is 1.7 um wide .

https://drive.google.com/file/d/1kSy_NjMnSB0UCjFtdO8S1lptvl1Jsl08/view?usp=sharing

Please see my file attached. 

[Guilin Sun]

 

It seems something is wrong in the settings, as you can see, the FDE result shows that neff is about 2.81 for the fundamental mode at 1.55um. However in the bandstructure result, there is no data at 1.55um. Before it reaches 1.5um it curves.

It might mix the modes as it has two bands.I checked the sweep. You set the relative kx from 0.25 to 0.7, which is too large. The kx in the sweep (and the group) is the relative one. Please check the First Brillouin zone, and refer to the example.

 

 

[minusunny]

 

 

 First Brillouin Zone  boundary is in pi/period = pi/0.222 = 14.15. So does my k  (= 2pi/period = 28.302 ) values range from 0.14 to 0.3?

 

 

[minusunny]

 

Please see my updated file:https://drive.google.com/file/d/1eDipYITeWEbXa-Y00AgC_A1P6HznivD2/view?usp=drive_link 

The effective index value still looks bad

[Guilin Sun]

I do not understand why you change the wavelength range now. Please check.

Now it has two bands. The other one may be real or fake, which will need you to identify.

The First Brillouin zone gives you the max k vectors at the symmetry points. In other periodic structure with square lattice, the relative kx is from 0 to 0.5. However now it is waveguide so its range will be smaller. 

Please visit this link  https://optics.ansys.com/hc/en-us/sections/360006919074-Photonic-crystals to check how to simulate the band structure (methodology).

 

[minusunny]

When i adjusted the kx for fundamental mode (0.25 to 0.45), the index i am getting is 2.08807 where as in the eigen solver of the source is 2.81, even if i adjust the values to a smaller range.. how do i get the effective index of first order mode ( the second TE mode from the user selct option in the source)? When i adjust the range  of kx in the sweep , i am getting the value of substrate effective index? should i use the same kx value or should i adjust the kx value in this case?

[Guilin Sun]

I did not find any guidance on how to select the k value. From what you mentioned it seems it is not the fundamental mode. try to use smaller k range. You may try to use broadband mode source for excitation, since the substrate does not support mode. I did some tests before and it works at least for some cases.

 

[minusunny]

Yes, it is the second TE mode. I am trying to get the beat length for a subwavelength grated structure. L pi = wavelength/(2*(neff0-neff1)). Is there any way to do this using banstructure simulation or should i use a different method?

[Guilin Sun]

You can use bandstructure to get effective index for sure. But you will need some understanding and scripting. bandstructure gives k (points not the real k) vs f, while for refractive index it is k0/k. you have to extract the data from band structure, it might not be every easy, since it is from 2D matrix from FDTD simulation.

Please make sure you simulate the mode you want.