Learning Forum

[Sangmyeong Yu]

I'm studying FDTD now but I have problem.

I want to achieve a structure and results similar to those in the mentioned paper, but my results are not matching. What should I do?

refence: Changhyun Lee et al "Near-flat top bandpass filter based on non-local resonance in a dielectric metasurface" Optics Express, 2023

[sagar]

I think you need to use the Periodic boundary condition to take the infinite array of the slab into account, instead of PML in x and y direction. PML can be used at the top or bottom of the unit cell. 

[Guilin Sun]

Thank you @Sagar for the reply.

 

In addition to the periodic BCs, please sweep the heights to get the transmission maps in the paper.

 

From your result, it is for one height, and the frequency points are only 5 I believe, then you will not be able to get results similar to the reference.

 

In addition you may need to read the example carefully to understand the settings and resuult.

Please also refer to this post as the comparison might not be very straight forward: 

Ansys Insight: Why my simulation result is different from published paper or experiment?

[Sangmyeong Yu]

 

At first of all, I tried to make 800nm~1000nm wavelength vs Transmission graph, but it’s not same paper’s graph.

(top) SiO2 thickness 160 nm,(middle) paper’s graph (bottom) SiO2 thickness 300 nm.

 

[Sangmyeong Yu]

It is my fdtd setting

[Guilin Sun]

The settings seems ok but some improvements should be done:

1: you will need to use override mesh to resolve the geometrically small features, with at least 4 mesh cells on each axis.

2: extend the geometry to outside of the Periodic BCs

It seems the geometry can be simplified to 2D if the structure does not change in the 3rd dimension.  But since this file is relatively small, it is fine to simulate in 3D.

Once you modified with the above 2 points, the next thing to consider is the material data they used. Make sure you use exactly the same as theirs. Otherwise I would not expect that you can get the same results. Fundamentally it is the materials that determine the result when the geometry is the same.