Learning Forum

[Handon Um]

When the pyramid size is reduced from 5 um to 0.5 um, the results of 1 um and 0.5 um pyramids show a special peak, which is larger than 1.

What is wrong with the setting of my FDTD?

[Amrita Pati]

Hi Handon,

I believe the issue is that the simulation reaches the end of simulation time before the auto shutoff level is reached. FDTD is a time-domain approach. The results are then converted into frequency domain by using the Fourier Transform. For the fourier transform to be accurate, it is important that the fields decay sufficiently at the end of the simulation. If not, you can obtain inaccurate results and a typical indication of this issue is the ripples you see in your spectrum. 

Please check the status of the simulation after you run the file.

If the status is 1, it means that the simulation ended before reaching the auto shutoff level. If that is the case, increase the simulation time by 10 times and rerun the simulation.

Regards,
Amrita

[Handon Um]

Thank your for your reply, Amrita.

 

I tried to increase the simulation time up to 100,000 fs (1um regular pyramid), but the result still showed the same peak as I saw before.

Is there other paramters in FDTD to resolve this issue?

 

Also, I wonder the resaon why the accuracy of FDTD decreases while the pyramid gets small. 

[Amrita Pati]

 

Hi Handon,

Thanks for sharing more information! From your screenshot, I can see that the simulation has still not reached the auto shutoff level as you can see that the status is 1. The status of the simulation should be 2 after you run the simulation and I can still see oscillations in the spectrum. I would suggest you to increase the simulation time even more. You can also right click on the auto shutoff result and visualize it to see what is its value when the simulation ends. 

Regards,
Amrita