Learning Forum

[Jingda Wu]

I am designing a ring resonator and using MODE to simulate the bending loss and neff. However, I find the result is very inconsistent depending on where the origin is set. 

 

In the original simulation of the lumerical example, the simulation center is at x=-0.1um, and a bending radius r=1.5um and here's the result

 

 

Then if I set the FDE origin to be x=0 and bending radius r=1.6um, The result is totally different.

 

 

This kind of inconsistency continues with user specified bend location. This makes it difficult to simulate complex structures. Is there a reason behind this inconsistency?

[Guilin Sun]

You are completely correct and it is expected. Please check this article: https://optics.ansys.com/hc/en-us/articles/360034917253-Solving-bent-waveguides-in-FDE-and-FEEM

Basically, it calculates the angular propagation constant and converts it to the linear porpagation constant:

while the angular "beta" is constant, the radius ca be arbirary. Therefore, you will need to correct the result using 

neff1*R1=neff2*R2

then the result will be constant when you change the bent center and bending radius. Please try and let me know if you have question.

[Jingda Wu]

I see. Thanks for the clarification.