Learning Forum

[Seyed Mohammadhossein Enjaviarsanjan]

Hello,

I am facing an issue related to defining a simulation time for measuring the quality factor of a high-Q cavity.

I employ a dipole cloud to excite the resonant modes within the cavity and, using a high-quality analyzer integrated into FDTD (Finite-Difference Time-Domain) simulations (you can find more information at https://optics.ansys.com/hc/en-us/articles/360041611774-Quality-factor-calculations-for-a-resonant-cavity), I attempt to determine the quality factor.

My question pertains to the simulation time. Given that my resonator possesses an exceptionally high-quality factor, ranging between 400,000 and 1.2 million, I encounter challenges in reaching the auto-shutoff level, for instance, 1e-5 or 1e-6.

Through several trials, I have observed that in some cases, a duration of 1000 femtoseconds (fs) is not as reliable, but there isn't a significant difference between 2000 and 1000 fs.

My assumption was that, since the method measures the slope of energy within the cavity and we are utilizing a time-domain approach to calculate the intrinsic quality factor, it should not be heavily dependent on the simulation time.

I would greatly appreciate your guidance on determining an adequate simulation time, especially since energy depletion does not occur within a reasonable timeframe. Additionally, if you could advise on what would be a reasonable auto-shutoff level, that would be immensely helpful.

Thank you for your assistance.

[Guilin Sun]

"My question pertains to the simulation time. Given that my resonator possesses an exceptionally high-quality factor, ranging between 400,000 and 1.2 million, I encounter challenges in reaching the auto-shutoff level, for instance, 1e-5 or 1e-6."

You should NOT persue reaching the auto-shutoff level, for instance, 1e-5 or 1e-6. Both the findresonance and FFT methods can give pretty accurate prediction from limited, relatively short time signals. Actually in order to get accurate result the signal should be large enough.

For the FFT method, you will need to make sure the signal strength in log scale  vs time be a perfect straight line. 

What is the best section of the time signal?  we should avoid the initial part of the source signal, which does not contain resonant information; and we also should to remove the later part, since its strength is low. Please check the script in the analysis group.