Learning Forum

[chrsa]

I want to calculate the powercouplinng coefficient for a ring coupled to a waveguide. As I want to simulate an anisotropically etched wavguide I choose a waevguide object similar to the ring modulator example. For medium rings (d = 300 µm left) the result seems to be more or less resonable (d = 300 µm, gap = 1µm Wavguide ~1.0 µm, lambda ~1.55), while for bigger rings (d = 900 µm, gap = 1µm Wavguide ~1.5´0 µm, lambda ~1.55) it looks, as if the ring consist of a wavguide with sharp bends simialar to the low resolution rendering. 

 

 

[Guilin Sun]

You can use ring object and modify the starting and stop angles to create the arc.

Bezier  https://optics.ansys.com/hc/en-us/articles/360034382234-Waveguide-Simulation-Object

https://optics.ansys.com/hc/en-us/articles/360034901673-Tips-for-creating-waveguide-bends-using-the-path-object 

Once the Bezier bend chooses the order, the number of points is fixed. and the curve should be smooth.

 

[chrsa]

Dear Guilin,

sure I could use a ring object for a ring, but sadly the ring object does not provide any non 90 degree sidewall angle. Regadring the Bezier - I also belive it should be smoth but somehow it is not for big objects.  

Do you have any other suggestions? I assume I could use a polygon and extrude it along a circular path. But is there any simpler solution than that?

Regards,

Christian

[Guilin Sun]

 

 

The Bézier-curve has analytical expressions once the spans are known.

You may try to explore from where it begins showing issue. You may file a feature request  at ix. https://ix.lumerical.com/

You may try to use other methods to create the side wall. For example layered ring, or  N-sided polygon toroid  in the structure library, or planar solid: https://optics.ansys.com/hc/en-us/articles/360034901573-Planar-solid-Simulation-Object