Excitation with Two Angled Plane Waves

fdtd

#1

Hello,

I have a question about excitation with two angled plane waves. I would be grateful if you could help me.

My simulation file has been attached. It is only free space, and I want to see the incident fields to be sure that I have a correct excitation for the future steps. I am using TFSF, and PMLs in all six directions. My plane waves are as follows: 1) propagating at 45 degrees between x and z direction. The E field is polarized at 45 degrees between -x and z directions. 2) propagating at 45 degrees between -x and z direction. The E field is polarized at 45 degrees between x and z directions.

The analytical result is easy (superposition of two plane waves). At x=0 plane, we expect to see Ex=0. I have written the following analytical solutions:
Ex = E02jsin(kx x) exp(-j kz z) / sqrt(2)
Ez = E02cos(kx x) exp(-j kz z) / sqrt(2)

My problem is that I cannot see this result. The following figure shows the real part of Ex at x=0 plane (monitor_1 in the file):

which is in contrary to the analytical solution that says Ex=0 at x=0 plane.

I would be so grateful if you could check my file and let me know how I can fix this problem. Generally, my simulation is a nano-hole in a metallic sheet illuminated by these two plane waves.

Thank you and best regards.

Test_incident_E_two_PWs.fsp (715.5 KB)


#2

Hi,
I had a look at your simulation file and as a general recommendation, you should leave at least one mesh cell between the grey area marking the source and PML boundary.

As for the results itself, it is not correct to expect Ex component to be zero. It is Ez that should be zero if we assume that both sources inject with the same phase. Of course, if there was 180 degrees phase shift, Ex would be zero and Ez would have amplitude of 1.4.

Here is result from 2D simulation(Ey is equivalent of Ez in the 3D simulation):

NOTE: There is a small bug when the position TFSF is offset by a half mesh cell. This means that if you want to land the field node exactly at x=0, the TFSF source position must be offset by dx/2:

This will be fixed in the future.

Here is the modified simulation file for 2D case:
Test_incident_E_two_PWs 2D.fsp (1.3 MB)