Learning Forum

[Saraah]

I have a photonic crystal, which I want to obtain the band structure for TE and the Tm mode. TE is the light propagating in the Y direction, and TM is the light propagating in the X direction.

I tried using random diple and specifying teh diple type as an electric dipole for Tm, and magnetic dipole for TE, but end up that the band structure I got is mixed, and no real difference between when I used magnetic or electric dipole. I also tried to change the angle of the (theta) in the dipole cloud, but I got the same thing. The boundary condition I used was Bloch for x, and y, and Z is PML.

This is what I want to see, TE and TM are denoted by solid and dashed curves:

[Guilin Sun]

Usually we have to use symmetry BCs to isolate the modes, please refer https://support.lumerical.com/hc/en-us/articles/360041567754-Photonic-crystal-cavity
Since your structure is 3D structure, and has substrate, it can be difficult to extract the specified modes, as either of the dipole types actually can excite other modes. A possible method is to first choose the proper dipole polarization, get the band structures. then choose a few typical frequencies in the bands, use a few monitors to get the Bloch mode profiles: Bloch mode profile - Photonic crystal
from the monitor results, you will know what the polatization of each. Since you know which band the frequency is located, then you can know polarization of the band.
So you will need to iteratively do the simulations.
If you want a clean final results, you will need post process to remove the undesired data points from the band results.

FDTD is time domain method so it is not like the PWE method which is easy to assign the polarization, unfortunately.

[Saraah]

I have a few things to get clear
'A possible method is to first choose the proper dipole polarization, get the band structures' ...... what do you mean by the proper dipole polarization choosing between magnetic or electrical dipole in the diople cloud?, and should I use a randomly printed dipole or angle dipole (theta)?
2. In the block mode example: how is the frequency is calculatedin (THz), and how can i get the exact number of that.
3.'from the monitor results, you will know what the polarization of each' . I didn't get this part, I have used 2 DFT monitors and did a couple of frequencies, and obtain the block mode for tye, but i don't know who to relate that to the polarization of my system.
4.In the Bloch example it says' Due to the symmetry of the problem, each of the bands calculated in the bandstructure above is doubly degenerate, one mode for each polarization. Thus care must be taken to excite only one of these Bloch modes. By placing a point dipole source on a symmetry plane of the lattice, one mode of a specific polarization may be excited directly.' Since my lattice has symmetry in X- a direction I place it in the x-direction, is this correct?

[Guilin Sun]

"'A possible method is to first choose the proper dipole polarization, get the band structures' ...... what do you mean by the proper dipole polarization choosing between magnetic or electrical dipole in the diople cloud?, and should I use a randomly printed dipole or angle dipole (theta)?
A: From my experience the dipole type does not affect the results very much. It is their polarization.
2. In the block mode example: how is the frequency is calculatedin (THz), and how can i get the exact number of that.
A: The frequency is chosen from the band structure. Please read the script to get knowing how the normalized frequency is calculated so you can get an idea. FOr example, in Planar3D_hex.lsf, it has
fs_all_norm = fs_all*a/c;
So the displayed normalized frequency if fs_aa_norm, and fs_all is the physical frequency.
3.'from the monitor results, you will know what the polarization of each' . I didn't get this part, I have used 2 DFT monitors and did a couple of frequencies, and obtain the block mode for tye, but i don't know who to relate that to the polarization of my system.
A: I know it is hard, as it is Bloch modes. But when you refer TE and TM, I believe you have your definition. Please compare the result with your definition.
4.In the Bloch example it says' Due to the symmetry of the problem, each of the bands calculated in the bandstructure above is doubly degenerate, one mode for each polarization. Thus care must be taken to excite only one of these Bloch modes. By placing a point dipole source on a symmetry plane of the lattice, one mode of a specific polarization may be excited directly.' Since my lattice has symmetry in X- a direction I place it in the x-direction, is this correct?
A: it says "maybe", so it is challenge but might be doable with a lot of try, and deep understanding of the Bloch modes as well as the TE TM polarization.

Since it is case-by-case, we can only give general guideline. Care must be taken to match the result with your theoretical definition. The simulated bloch mode has all the E and H field components ready to be analyzed.