Ring resonator grating and "engine error" problems

grating
whisperinggallery

#1

Hello!

I would really appreciate your help.

I am trying to simulate a ring resonator with out-coupled gratings on top, and excuse me for being new to Lumerical, but I have no idea how to make the gratings to bend about the ring, maintaining a position in the centre of the ring.

I tried to do this with just a few small gratings, and after adding sources and the like, I still get this error:

I defined my own materials (as n,k materials), and here is the view of the gratings (I just put a few since I got tired of copy pasting and wanted to find a better, more automated way of getting the gratings about the ring):

I am at a loss…no idea how to approach this problem. I would GREATLY appreciate any help.

Thank you so much.


#2

Usually the best way to draw gratings or any periodic structures is using the script…
This is an example of how you can create your one structure group and define its properties through the script:
https://kb.lumerical.com/en/pic_passive_grating_coupler_3d.html

If you are new to scripting, please check the knowledge base scripting section before you start.

We would certainly like to help you. So, what you want is thin rectangles rotated such that they form a circle of the same center as the ring resonator. Please, confirm. In this case, I can upload a ready made structure for you.


#3

Dear Aya,

I thank you so much for your reply.

I appreciate so much. Yes, I confirm this is what I like. I have no idea how scripting works, I hope you can teach via your example.

Thank you!


#4

Dear hdh,
Check this post please and tell me your feedback.
You are welcome.


#5

Dear Aya,

Thank you so much. I really appreciate. I have a lot to learn and I hope to learn more from you (I am a graduate student at Northwestern University, by the way working on photonics).

  1. How did you create you script? Any way I can learn the general language?
  2. Attached is my fsp file. Whenever I run, I get “engine error” problem. Why?

Thank you for your time.
ring_QCL_np.fsp (1.8 MB)


#6

Hi, to check the engine error, right click on job error, view job details:
Here is what I got when I did so. The problem is insufficient memory.


Try the same simulation for smaller dimensions first. Then,

  • To solve this problem try using symmetric boundary conditions.

  • Another point is why is the plane wave source smaller than the structure? Also, only a small part of the rings lies inside the simulation domain so the ring is basically unseen?

You can send me the paper/reference you are trying to mimic to better understand what you need to do.


#7

Thank you Aya.

Basically, this is a ring quantum cascade laser (QCL).

The gain medium (labelled core) provides the light that is then supplied to the ring waveguide, and the purpose of the ring gratings (second order grating) is to outcouple this light. I just want to find a way to do this: get the field profile of the output light from the waveguide.

There are many papers regarding ring quantum cascade lasers. I can send them if you like. I just wanted to do a general simulation.


#8

By the way, I am also using trial version. Maybe that has something to do with the engine error?

Even when I completely disable the ring grating, it will not work. Right-clicking to check job details gives me very little information:

Notice I put the source in the “core” region, since that is the gain medium of the QCL, even though I am not incorporating gain in this model (assuming laser is just below threshold). The QCL is intrinsically TM polarised, so I have the injection-axis as the z-axis.

I am sure the error I am getting is because I am setting everything up incorrectly.


#10

Hi, I am unable to read the error message since the text is too small and the image can’t be expanded, but as Aya previously suggested, the problem is most likely related to the memory requirements.

As you can check on your own by going to Main toolbar->Check->Requirements, the simulation requires over 2000GB of RAM:

Such high requirements are mostly due to the size of your 3D simulation that is over hundred microns. This is very large for FDTD.

Also, I looked at your simulation and I would point out the following:

  1. The source is placed above the structure and injects into PML, meaning that the light will not interact with the structure.

  2. Using PML alongside the direction of propagation with plane wave source will result in edge effect as described here:
    https://kb.lumerical.com/en/index.html?ref_sim_obj_planewave_edge.html

  3. Only portion of your very large structure is in the simulation region. This means that most of the structure will not be considered. Also, the monitor is placed outside of the simulation region and no data will be recorded.

Since you are new to Lumerical, I would recommend to go through the getting started examples that explain step-by-step how to correctly setup a simulation and get reliable results:
https://kb.lumerical.com/en/index.html?getting_started.html

If you are interested to learn Lumerical scripting language, the following tutorials should help:
https://kb.lumerical.com/en/index.html?getting_started.html


#11

Sorry, I don’t understand how “only portion of [the] very large structure is in the simulation region.”

The picture shows monitor and source and FDTD region (orange) completely covering the ring.

I will try the getting started examples, but in the link you send me, there are tutorials, which contain the files for the programs, but where are the steps? For example, in https://kb.lumerical.com/en/index.html?solvers_ray-vs-wave-simple-example.html I downloaded the fsp file and still got an engine error, even with such low necessary computation requirements:

Job manager doesn’t seem to be of much help either–is this all because I have the trial version?


#12

Q: Sorry, I don’t understand how "only portion of [the] very large structure is in the simulation region.

It seems that the file you have shared with us is different than the one from your screenshots and only portion of the structure is in the simulation region:

Q: I will try the getting started examples, but in the link you send me, there are tutorials, which contain the files for the programs, but where are the steps? For example, in https://kb.lumerical.com/en/index.html?solvers_ray-vs-wave-simple-example.html

Only the “Getting started” examples on this page have step-by-step guide. You might be particularly interested in this 3D FDTD ring resonator example.

Q: I downloaded the fsp file and still got an engine error, even with such low necessary computation requirements:

This is a good test test. The trial license is not limited in functionality. It seems like it could be a problem with mpi configuration. Particularly the Yen character in the file path seems suspicious.
Try to go to you resource settings:

and bypass the mpi in the advanced settings as shown here:

Note that when the mpi is bypassed, the performance might be poorer. Therefore, it is more of a test to try if the mpi configuration is the source of the problem rather than a solution. A possible solution could be to modify the path on the same tab so it does not include the Yen sign as shown on the screenshot above.


#13

Thank you for your reply.

On the topic of the msi issue, I tried bypassing it, but am still running into the same issue. My computer is Japanese, so all the “” delimiters are switched to ¥ by default.

I think you are right–this is the problem. I do not know how to fix it. Should I install the Japanese version of Lumerical perhaps?


#14

We have seen another case that had issues with the license due to a Japanese character on the computer name used by the machine running the license manager. We can try to use the Japanese version of the Software, but we can also try to add the English keyboard input (if its not your default keyboard input) into Windows if this will help with the Japanese character.