Learning Forum

[Osaka]

In a periodic structure, does the value of the cross section change depending on the position of the monitor and TFSF source?

[Guilin Sun]

Since TFSF is sensitive to mesh, so the result may change with TFSF and monitor locations. IN general, we suggest to use uniform mesh for TFSF region, at least in the directions perpendicular to the injection plane normal. In addition, since scatting can have very large angle, the PML should be sufficiently far way from the objects, and thick to absorb efficiently the scattering fields.

[Osaka]

Thank you for answering the questions! I'm sorry I didn't explain condition enough.
I want to calculation value of cross section of periodic structure, so I set the boundary condition as anti-symmetric in the x-direction, symmetric in the y-direction, and PML in the z-direction to create periodic structure. Also, I use uniform mesh for TFSF region.
If I set the boundary condition as PML in the x, y and z direction to create one particle, value of cross section doesn't change depending on the position of the monitor and TFSF source. However, If I set the boundary condition as anti-symmetric in the x-direction, symmetric in the y-direction, and PML in the z-direction to create periodic structure, value of cross section change depending on the position of the monitor and TFSF source. Does this means that I can't calculation cross section value in periodic structure? and Why?

[Guilin Sun]

Traditionally cross section is only applied for isolated object. It is recently that people are interested in periodic structures, which are grating-like diffraction orders in the far field. Based on the same principle you can calculate the cross section of periodic structures. However you only need two monitors in z. In such case, the resonance can be strong thus it needs longer simulation time and smaller autoshutoff min. Because of the plane wave nature, the TFSF must be larger than the XY spans of the simulation. eg, it has only two planes perpendicular to z that are effective. If the TFSF's xy span is smaller than periodicity, sure the result will change since you are inject partial plane waves to the periodic structure. Please make sure the xy spans of TFSF is larger than those of simulation region. Similarly for the monitors.