Mesh and Simulation Region Sizing


Dear Lumerical Community,

I am wondering if my current mesh and simulation region sizes are properly set. I have created a cavity consisting of 4 beveled PEC plates with a TE polarized electric dipole source and am attempting to measure the power on the side of the cavity where the dipole source is located and compare it to the power on the opposite side of the cavity. I am trying to understand how large my mesh and simulation regions need to be in order to obtain the most accurate results.

Does my simulation region need to completely enclose the mesh region within it, or vice versa?

In conjunction, should I encompass my entire cavity structure in both the mesh and simulation regions?

Are the PML boundaries at the edges of the simulation region?

Attached is my current simulation file. What, if any, changes would one make within my file to ensure that the results I am obtaining are as accurate as possible?

Fig_2a_Ag_Plates_Dipole_1mm_Outside_No_Apod_PEC_Variable.fsp (284.2 KB)


Hi @cnorvill
I found in Lumerical’s knowledge base something vey useful about point sources(dipole). Here is the link: and also you can download the example: “usr_source_electric_dipole.fsp” .
I compared carefully your file and Lumerical’s example and i found many differences. I would recommend that you should use 3D-FDTD simulation. The simulation region should encompass the four beveled PEC plates and the PML boundary conditions should end before the end of the plates. Another advice is to use mesh type: auto non-uniform.
In boundary conditions tab—>PML settings—>profile : standard---->layers: 8
I hope that it helps.