Rotating a 2D grating


If I use a 2D Grating Coupler component on my model and rotate it, am I to understand that it only applies to a 2D space or does it expand across the 3D surface?

I have a plane surface (xy) with a grating structure built up in the z direction in 2D. The alignment of the grating rotated to some angle between x and y. The structure is exposed to light polarized in the y direction, normal to z, to read the transmission through the structure in x polarization.

Is this clearly worded?


Hi @t.blackwood.1

Thanks for reaching out.

If you are performing a 2D simulation, software will ignore the third dimension and only considers the cross section where you have defined the FDTD region (and assumes that the cross section continues to infinity!). If you can attach your file and explain what you are looking for, I can be a better help.

Please let me know if I could answer your question.

Rotating 2D Grating Couplers

@bkhanaliloo, thank you for replying!

3period_CustomGrating_auto.fsp (499.9 KB)

I’ve managed to create something of what I was looking for by setting the grating period and angle, ungrouping it, giving it a third dimension, and then regrouping it. Though I am trying to look at it in the visible range, I am currently thinking I have to look at the meshing. Current simulation time is (In Austin Powers: Doctor Evil voice) One Million Hours… Well, ~100,000 hours…

Any further advice is greatly welcomed!


Dear @t.blackwood.1

Thank you for reaching out.

Is there any reason that you want to start with such a fine mesh specially on the z-direction (in your file it is set to 1e-44 :confused:)? Definitely this will impact your simulation time and memory drastically.


Hang on… (…dz = 1 e -44 um…?) Excuse me whilst I introduce my forehead to my desk.

This was originally a 2D file that I adapted from another researcher’s work. I’m sorry to bother you with my silliness!

I’ve set it to 0.005 um, which I’m guessing should work well with my 400-700 nm band and materials. Unless your experience would recommend a better resolution?

The sim time has reduced to 4 hours 30 mins, bit more acceptable!


Hi @t.blackwood.1

I hope it didn’t hurt :slight_smile:

You can start with a coarse mesh all the time, please see our KB example on grating coupler, and you should find a lot of useful information about proper meshing such as here. Our KB examples should be a great starting point for you.


Ach, it’ll get better!

Thank you for the links! Managed to produce a result! Might be incorrect result but it is a result! Round 2 now interpolates the materials index w.r.t. wavelengths…