2D FDE mode solver


Hi everyone
I am constructing a waveguide in mode solution using 2D FDE solver to support 3 modes. The problem is when putting this in FDTD solution and adding a mode source to enter one mode, I do not get the modes I saw in mode solution to choose from them!! I got new modes with different refractive indices (the effective index is important for the whole design parameters, length, coupling,…).
Any help will be appreciated.

Getting started in Lumerical Mode solutions

If you have exactly the same structure, material and meshing map in both solvers, it should be possible to find the same mode in the Integrated Mode Source and FDE since the Mode source basically uses te FDE engine to find modes. One way to ensure that is to copy and paste(ctlr +c &v) the simulation objects from FDE to FDTD. If you still having trouble finding the modes you want in FDTD, please reply this post again with the your files, if possible.


Thank you for your reply. I did find them if I am using 1D FDE mode solver in the mode solution. If I am using 2D FDE mode solver, I get different modes.
The solutions from 1D and 2D FDE are different !


I am not sure if I understand your reply.

If this is a 1D problem, then I think it is reasonable to expect the same solution from 1D and 2D solver. If this is a 2D problem, then I probably will not recommend 1D solver. If you dont mind, can you elaborate more how is this related to the Integrated Mode Source in FDTD? It will be helpful if you can upload your files here too!


waveguide_multimode_FDTD.lsf (4.0 KB)
waveguide_multimode_mode.lsf (2.8 KB)

I uploaded the script file to get the modes using 1D FDE and 2 D FDE mode solver.

You can run ‘waveguide_multimode_mode.lsf’ in the mode solution
In the case of 1D, here is the results from mode list. I showed the first and second modes for comparison.

In the case of 2D, here is the results for the first and second modes

you can notice the large difference in the effective index values. This can be understood as the 1 D solver sees the waveguide as a slab not 2d strip waveguide. So the second case is the actual one that I want to use in FDTD simulation.

when running the script ‘waveguide_multimode_FDTD.lsf’ in the FDTD solution and open the edit window for the mode source, here is what I get

It is not the 2D solution I want !! I hope I can make this clear.


Hi. As @kchow mentioned, when you want to compare your modes from 1D and 2D FDE solvers, your structure should be 1D because otherwise you can not use 1D FDE. Now looking at your structure from the “waveguide_multimode_mode.lsf” file, it is clearly a 2D structure.

When you use a 1D FDE to look at the mode of this structure, the equivalent structure in 2D will be very different (see below):

Now if you compare the modes between these two cases, i.e. the 1D FDE on left to the 2D FDE on right with the waveguide uniform along z-axis (Please note that I just used materials from the library since I did not have access to your “materials” script):

First TE mode:

Second TE mode:

Note that the index values are slightly off since in the 1D case the waveguide in infinitely long along z whereas in the 2D case it gets cut off at the +/- z boundaries. For the same reason the ordering of the modes are different as well.

NOTE: Looking at your structure, I would recommend using 2D FDE since 1D FDE is giving you modes for a completely different structure. Similarly, you would need a 3D FDTD to simulate propagation in the structure.

NOTE: Since your structure has no variation along x, I don’t think you need to do a simulation in FDTD. Even if you have some variation eventually (like a taper), you would probably be better off using the EME solver in MODE.


Thank you very much. It was of great help.

I am studying a directional coupler and how a mode will couple to a second waveguide and this was part of the problem. So I guess eventually I will need to do 3D FDTD simulation. won’t I??


Hi. There will definitely be exceptions but in most cases my understanding is that you should be able to simulate coupling between multiple waveguides using the EME solver in MODE Solutions. Of course a 3D FDTD simulation will be able to do the same thing but the EME solver should give you the same result much faster. To see some examples of waveguide coupler simulations using the EME solver check out this page in our KB: Waveguide Couplers.


How can I get a figure like the right one in the mode solutions? I found I can only get the left figure. Thank you!


As @aalam pointed out in his post, the left picture is showing 1D FDE solver results while the right figure shows 2D FDE results. Hence, 1D eigensolver can show E field intensity on the y axis while 2D is using color bar to show the intensity distribution since the y axis is used for spatial distribution in the z direction.


Thank you very much for your answer. But excuse me, I have another question about how to run the script files “waveguide_multimode_FDTD.lsf” and “waveguide_multimode_mode.lsf” which @omnia.nawwar uploaded. I found there always exists an error-materials is not a valid function or variable name when I run them in mode solutions. So what should I do to run the script files correctly?


Hi @hydzp,

It seems “materials” is a scripts called by those 2 scripts to create/add the materials used in the simulations. You would need that script as well or replace these materials by ones available in the material database.


Oh, that is fine. Thank you very much for your answer.