why the source is set so long compared with the waveguide core in the line defect example?



I’m repeating this example (Line defect) from knowledge base, why the source is set so long compared with the waveguide core in the line defect example?


The source here is a mode source (see this KB page for details) that calculates the optical modes in the waveguide. By default it uses the fundamental mode as input signal but the user can select other modes as well. The source works in a way similar to the FDE solver in MODE Solutions to calculate the eigenmodes. The extent of the source defines the extent of the solver region for the eigenmode calculation. So just like the FDE solver, the extent of the source needs to be much larger than the waveguide to ensure that the field intensities go to zero (or very small value) at the boundary.

NOTE: The mode source uses PEC boundary condition at the boundaries to calculate the eigenmodes.

NOTE: If the extent of the source gets truncated by the solver region boundary then the solver boundary sets the extent of the source in that direction. However, the boundary condition used by the source is always PEC and is not set by the boundary conditions of the solver region.


Thank you for your detailed answers to my question! As I 'm strating to use FDTD solutions, but I don’t find the example to calculate transmission and reflection of waveguide in FDTD solutions, the example above is in MODE solutions, so I have several questions about waveguide simulation in FDTD solutions:
(1)If I want to simulate the waveguide in FDTD solutions, is it right to set the simulation the same as in the example in MODE solutions? The attachment is my attempt to do this, however, the result is not good.
(2)Is it better to use MODE solutions to simulate waveguide than FDTD solutions? If it is, could you please recommend some tutorials about MODE solutions?
Thanks a lot!
line defect of square lattice.fsp (355.3 KB)


For waveguide tutorial in FDTD and varFDTD, I would recommend this example to look at. varFDTD is good for planar structure where there is no vertical coupling between slab modes. Here is a handy flow chart to help you choose which solver for your structure.

I am guessing your simulation file may be related to this post. If this is the case, I think it will be better for us to continue the conversation in that post.