Learning Forum

[Luise Armbruster]

We want to use the EME in order to investigate the propagation of the guides modes through a structure. Unfortunately, finding all guided modes in each slice of the structure required searching for 100 modes. Hence, the simulation requires an enormous amount of memory. Since we know the number of guided modes at the beginning of the (tapering) structure, which is far below 100, we would expect much less modes to suffice for the description of the total field in each slice. Inspecting the computed modes more closely, we noticed that the EME eigensolver does not seem to search first for the modes with an effective index closest to the highest refractive index value. Now, the question is, if there is any possibility to change the settings, so that Lumerical searches for the modes near a user-defined refractive index first.

[Kyle]

Hello,

You can change the index where the mode search begins using the custom eigensolver settings:

 

Using symmetry boundary conditions may also help to reduce the number of modes required.

[Luise Armbruster]

We made the settings as described and also got the guided modes in the "EME Solver Cell Groups Setting" window when we let 10 modes be calculated. But when we ran the simulation, we didn't even get all the guided modes for the first cell.
What could be the cause for this?