the question about the polarization rotator


dear all
recently I have learnt polarization rotator example on your official website.But I have some questions want to ask. In the taper_width_sweep.lsf

store top 5 modes for largest waveguide width

since the mode order may change for different width

if (i==1){
how can you define mode1 is TE0,and mode2 is TE1,mode3 is TE2,mode4 is TE3 ? I dont know how to get it.
2. slab_neff = 2.754047; # from a reference simulation without the ridge
How to get the slab_neff value?
3.In the reference paper I have get the calculated effective indices for the eigen modes of SOI rib waveguide

it looks a little different from the results you get(as shown below)

do you have any opinion about it?
Hope to get your feedback.Thanks in advance.taper_width_sweep.lsf (1.4 KB)
Mode_conversion_in_tapered_submicron_silicon_ridge.pdf (3.5 MB)
taper_width_sweep.lms (238.0 KB)


Dear @jbwei

If you look at the script, we change the ridge width from 3um down to 0.5um. If you manually set the ridge width to 3um and run the simulations, here the first 5 modes that you will get:

As you can see the first 4 modes are TE and the last one is TM. Without looking at these modes, there is no way to know the order of the modes as it changes with geometry and wavelength.

  1. If you disable ridge (as is explained in the script) and run the simulation, you can find the slab_neff:

  1. At ridge width of around ~1um, we have a mode crossing between TE1 and TM0. In taper_width _sweep, we use bestoverlap command to track the modes and I think our simulation results are correct. To check the mode crossing, I also set the ridge width to 0.8um and then 1.2 um, and ran the simulation for each. You can clearly see that the location of TM and TE1 are reversed which means that the effective index of TM0 is greater (smaller) than TE1 at ridge width of 0.8um (1.2 um).

I hope this answered all of your questions.


dear bkhanaliloo
thanks for your help.I think you are right. Now I will go on learning more about the polarization rotator. Thanks so much.


I am glad that it was helpful :slight_smile: