EME Optical Fiber Mode Divergence

[Wender Daniel]

Hello, I want to simulate the divergence of the mode after exiting a optical fiber usign EME solver. How should I set the EME cells to get this mode expansion?  

[Guilin Sun]

If I am correct, you want to simulate the divergence angle of a fiber mode in EME? 

Ideally, you can either FDE or FDTD to get the farfield pattern. FDTD might be more intuitive. Once you get the farfield, you can refer to this exmaple to get the divergence angle: Creating line plots of far field projections

[Wender Daniel]

Hi Guilin Sun,

I am simulating an optical fiber edge coupled to an silicon inverse taper in EME. I want to get the increase in the coupling loss as a function of the distance of the fiber to the taper. I think that the mode divergence must be considered. How can I set the cell between the fiber and the taper to consider the increase in the mode size?

 

Thank you,

[Guilin Sun]

Do you mean you want to simulate the result with varying the distance between the input fiber and the edge coupler, where the gap is in air or other material without light confinement? In this case the gap part will need many modes to build up the fields accurately. you may also consider using import sourece for EME, with FDTD simulation of a fiber and air, and use farfieldexact to calculate the fields at any distance. The calcualted EH fields then can be imported into EME. Both methods work but need some time to experiment for a better results.

 

[Wender Daniel]

Yes, that's exactly what I mean. How many modes in the EME cell are reasonable? I have tested up to 100 modes but didn't get the mode divergence.

[Guilin Sun]

Since the air gap does not support regular modes, you need to test the numbers as the fields are continuous there. When the gap is small, the beam may not change significantly. Such simulation to obtain the relationship between the coupling and gap distance is tedious and needs a lot of time to try.

[Wender Daniel]

Hi Guilin,

I have another question.

I'm getting unphysical modes in the edge coupler cells. I wondering if these modes are leading me to incorrect results. Should I remove these modes from the simulation?

 

Thank you,

[Guilin Sun]

The so-called unphysical modes are not propagating modes I guess. Instead they are radiation or leaky modes to help fulfilling field-continuity of the tangential components, which is the key principle of the EME method: the tangential fields are continuous.

[Wender Daniel]

Yes, the modes are note propagation modes. It seems like modes created by the metalic boundaries. I only get results that make sense when I remove these modes from the simulation.

 

[Guilin Sun]

Usually those modes are needed to have continuous tangential field. In some cases such modes contribute very little to the final results. You can play around to find the best condition for accurate result.