Farfield projection from single monitor versus box of monitors method

fdtd
farfield

#1

I am attempting to get the far field projection of a dipole that has part of its emission being redirected by a metal lens reflector.

I started using the file to understand the far field projections. I see there is the single monitor method and the box of monitors method. I was able to reproduce the analytical expected results from the dipole in the homogeneous environment using the box of monitors.
When I try the single monitor method, the result does not quite match up as expected. Please see the attached far field polar plots from the XY halfplane below the surface. Filename: basic_test_lumerical_monitor_farfield.jpg is for the single monitor farfield in the XY_halfplane
Filename: basic_test_lumeical_box_farfield.jpg is using the analysis group for far field box scat_ff and the usr_farfield_symmetry script supplied on the website.

Is this due to simulation size, resolution, or mesh size? Or that the single monitor is too far away from the source so that the field is absorbed by pml before reaching the monitor? I tried placing a far field monitor closer to the dipole and the resulting far field was different. The reason I want to trust the single monitor is because of the next part of the work.

My second set of questions is for the simulation with the lens. Can I use the box method here, or will it give incorrect results. Currently the box method and the single monitor method do not agree here either. Please see the attached simulation file: x-dipole_orientation_lens_oxidesimulation_test

While I believe the single monitor method is correct in this case, because the medium containing the monitors is homogeneous, I assumed that that box method should work as well.

Thank you,

Brandon Demoryusr_farfield_symmetry.fsp (4.3 MB)x-dipole_orientation_lens_oxidesimulation_test_symmetryboundaryconditions_v3.fsp (413.7 KB)


#2

Hi,

A quick answer is that you will likely need a single planar monitor for the projection. The box method is only for projection where all monitors are in homogeneous medium. This KX post and KB article have more info


https://kb.lumerical.com/en/layout_analysis_projections_from_monitor.html

Regarding the simple test, I think you will need the farfieldfilter before the far field projection. If you still have problem with that, please reply this post.

https://kb.lumerical.com/en/index.html?solvers_far_field_projections_far_field_filtering.html


#3

So I understand that the box should be used in the homogeneous medium. What I am wondering about is if you have the reflector outside of the box, will the reflector distort the results when compared to a single monitor? I have looked at the results from the single monitor and the box monitor set. They are different for the far field. But the box is placed in a homogeneous environment.
My thought process is that the box is projecting the components in each direction and adding them. But if there is a reflector outside of one side of the box, the far field does not make sense. Therefore, the box method would not work.

I will stick with the single monitor. I just needed this as a sanity check. Thank you for the information.


#4

I have to admit that I did not 100% understand your questions, especially the part involving a reflector outside. If there is a reflector outside, I don’t immediately how do you make sense of the far field. The far field projection assumes the same homogeneous medium from near to far. If there is reflector, I think this is breaking this assumption. If you can show us some illustrations to help us understand the design in your mind, I can try to provide more comments.