I am wondering if you could let me know about the group velocity of the waveguide bend. Is this the same as calculating the group velocity for a circular ring resonator of this radius? I would like to calculate the intrinsic Q for ring resonators, and it seems like cylindrical geometry 2D mode solver would be the best for this. But I would like to make sure that I am getting the correct ng, the number I get seems smaller than I would expect.

# Calculation intrinsic Q with waveguide bend

**aya_zaki**#2

This can be done on MODE solutions. If you have the group index of the bent waveguide and the imaginary part of the effective index, the intrinsic quality factor can be calculated as

Q= Ng/2*n_eff_im

So here I plotted the Ng of a 1.5 um bent Si waveguide:

Since the Imaginary part = 8.35e-5, the quality factor calculated would be 26,000. According to this paper [Silicon microring resonators with 1.5-micron radius.pdf (444.2 KB)] the intrinsic quality factor based on 3D FDTD simulation is about 50,000.

The bending loss estimated by MODE solver is 29 dB/cm (0.0029 dB/um) at wavelength = 1.55 um. The attached paper shows bent loss of about 8 dB/cm for the 1.5 um radius in figure 1.

Hope this helps!

Thanks for your response. You wrote "The attached paper shows bent loss of about 8 dB/cm for the 1.5 um radius in figure 1”, which is very different to the bending loss you simulated (29 dB/cm). Why do you think there is such a big difference?

**esmagden**#4

@aya_zaki is there a reason why in many simulations, Im{n_eff} comes out negative? I’m guessing it has to do with which boundaries are chosen to be PML vs metal, but are there any rules to follow on this?

And what options did you use to get the said “Imaginary part = 8.35e-5” in terms of simulation boundaries, PML layers etc? Can you attach the file you used here?

Thanks.