Note: This discussion is about an older version of the COMSOL Multiphysics® software. The information provided may be out of date.

Discussion Closed This discussion was created more than 6 months ago and has been closed. To start a new discussion with a link back to this one, click here.

Fiber Model in 3D Analysis

Please login with a confirmed email address before reporting spam

Assalamualaikum. I am Redwan Ahmad. I am trying to analyse a fiber model in 3D format in Comsol Multiphysics 4.4 for investigation of longitudinal properties. But I am facing some problem, as follow as:
1) In Electromagnetics waves, frequency>> Scattering Boundary condition>>…. Which region I have to select? (Only Cladding)
2) In Electromagnetics waves, frequency>> Port 1 & Port 2>>, may I have to select core region?
3) In Study>> Frequency Domain>> there is an option about Frequency and laod parameter values. Can you please explain about it?



3 Replies Last Post Dec 18, 2014, 8:24 a.m. EST

Please login with a confirmed email address before reporting spam

Posted: 9 years ago Oct 20, 2014, 9:41 a.m. EDT
Dear Redwan,

I'll try to answer one by one your questions:

Question 3 , taken from the documentation:

"Specify the frequencies to use for the frequency sweep. Type the frequencies in the Frequencies field using space-separated numbers or the range function.
Use the Load parameter values field to select a file with parameter values. You can browse your file system for files by clicking the Browse button. After selecting a file click the Read File button to load the parameter values into the Frequencies field."

Question 2:

Select the core and the clad region at the input and output face. Also you have to do one boundary analysis for each port. In the model you have uploaded the Scattering Boundary Conditions are at the input and output faces. To simplify your model I suggest you erase the Scattering Boundary Conditions, at least for now. And put instead of them the ports at each input/output face of the fiber.

Now a question for you:

What are you expecting to get from this model? Is it the refractive index? Then you dont need to do a propagation study to have the effective refractive index of an axially symmetric fiber. Try a modal analysis.

About your model:

Basically, your main error is trying to do a time domain study using a FEM software. The finite element method is suited for eigenvalue and time-harmonic problems. Comsol has a workaround for this limitation and you can use it if you have the Wave Propagation Module. Anyway, dont expect to propagate a huge extension (many wavelengths) using this module.

If you are looking for a propagation method, then I suggest to look for FDTD or BPM.

Another problem was defining the ports. If you want to propagate the fundamental mode, you need 2 ports at each end of the fiber (one for each polarization). Also you need to specify the boundaries for each mode (input and output end).

In the geometry you selected the axis (x,y,z) = (1,1,1). I have changed it to (x,y,z) = (0,0,1) in that way you only have to work with one initial component (z).
Also I'd reduced the whole structure in order to reduce the number of modes and simplify the analysis.

Also in geometry, you selected "Layers on side" for both cylinders. I didnt get the use of it, so I'd unchecked it.

The frequency you used for the modal analysis is not equal to the frequency used for the frequency domain study step. For that reason, I've defined f0=c_const/1.55[um] as a global parameter and used it in both study steps.

In order to learn how to do the modal analysis, look in the "Directional Coupler" tutorial.

Another advice: when using FEM you have to avoid using 3D models as much as possible. For this model you may have used a modal approach (longitudinal symmetry) or using an axially symmetric model (angular symmetry).

Notice that the attached model shall not converge to a solution, since you will need lots of memory to propagate using plain FEM. Since I dont know if you have the Optics Module, I'd prefered not to include it in the attached model. This model has most of your errors corrected. From there and including a Beam Envelope study (under the Wave Optics module) you may be half way to success.

For a detailed guide of how to model a waveguide using the Wave Optics module, look for the directional coupler tutorial.

Hope these helps.

cheers,
Felipe Beltran-Mejia
Dear Redwan, I'll try to answer one by one your questions: Question 3 , taken from the documentation: "Specify the frequencies to use for the frequency sweep. Type the frequencies in the Frequencies field using space-separated numbers or the range function. Use the Load parameter values field to select a file with parameter values. You can browse your file system for files by clicking the Browse button. After selecting a file click the Read File button to load the parameter values into the Frequencies field." Question 2: Select the core and the clad region at the input and output face. Also you have to do one boundary analysis for each port. In the model you have uploaded the Scattering Boundary Conditions are at the input and output faces. To simplify your model I suggest you erase the Scattering Boundary Conditions, at least for now. And put instead of them the ports at each input/output face of the fiber. Now a question for you: What are you expecting to get from this model? Is it the refractive index? Then you dont need to do a propagation study to have the effective refractive index of an axially symmetric fiber. Try a modal analysis. About your model: Basically, your main error is trying to do a time domain study using a FEM software. The finite element method is suited for eigenvalue and time-harmonic problems. Comsol has a workaround for this limitation and you can use it if you have the Wave Propagation Module. Anyway, dont expect to propagate a huge extension (many wavelengths) using this module. If you are looking for a propagation method, then I suggest to look for FDTD or BPM. Another problem was defining the ports. If you want to propagate the fundamental mode, you need 2 ports at each end of the fiber (one for each polarization). Also you need to specify the boundaries for each mode (input and output end). In the geometry you selected the axis (x,y,z) = (1,1,1). I have changed it to (x,y,z) = (0,0,1) in that way you only have to work with one initial component (z). Also I'd reduced the whole structure in order to reduce the number of modes and simplify the analysis. Also in geometry, you selected "Layers on side" for both cylinders. I didnt get the use of it, so I'd unchecked it. The frequency you used for the modal analysis is not equal to the frequency used for the frequency domain study step. For that reason, I've defined f0=c_const/1.55[um] as a global parameter and used it in both study steps. In order to learn how to do the modal analysis, look in the "Directional Coupler" tutorial. Another advice: when using FEM you have to avoid using 3D models as much as possible. For this model you may have used a modal approach (longitudinal symmetry) or using an axially symmetric model (angular symmetry). Notice that the attached model shall not converge to a solution, since you will need lots of memory to propagate using plain FEM. Since I dont know if you have the Optics Module, I'd prefered not to include it in the attached model. This model has most of your errors corrected. From there and including a Beam Envelope study (under the Wave Optics module) you may be half way to success. For a detailed guide of how to model a waveguide using the Wave Optics module, look for the directional coupler tutorial. Hope these helps. cheers, Felipe Beltran-Mejia


Please login with a confirmed email address before reporting spam

Posted: 9 years ago Oct 22, 2014, 1:44 p.m. EDT
Dear Felipe Beltran-Mejia,
Thank you very much for your reply. It's really helped me very much. Wish you good luck.

Best Regards
Redwan
Dear Felipe Beltran-Mejia, Thank you very much for your reply. It's really helped me very much. Wish you good luck. Best Regards Redwan

Please login with a confirmed email address before reporting spam

Posted: 9 years ago Dec 18, 2014, 8:24 a.m. EST
Dear Redwan,

since this is a recurrent question I find on the forum, I had posted a model that shows how to model an optical fiber in 3D. Please find it in the Model Exchange section:

www.comsol.com/community/exchange/251/

As you will see, it is a directional coupler. If you want to model a single core fiber, all you have to do is delete one of the cores and compute again the effective indexes for the Boundary Mode Analyses before computing the whole study.

hope it helps,
Felipe Beltran-Mejia
Dear Redwan, since this is a recurrent question I find on the forum, I had posted a model that shows how to model an optical fiber in 3D. Please find it in the Model Exchange section: http://www.comsol.com/community/exchange/251/ As you will see, it is a directional coupler. If you want to model a single core fiber, all you have to do is delete one of the cores and compute again the effective indexes for the Boundary Mode Analyses before computing the whole study. hope it helps, Felipe Beltran-Mejia

Note that while COMSOL employees may participate in the discussion forum, COMSOL® software users who are on-subscription should submit their questions via the Support Center for a more comprehensive response from the Technical Support team.