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 The information provided may be out of date.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.

effective index of a simple index-antiguided fiber

Please login with a confirmed email address before reporting spam

Hi,

I want to get the effective index Neff of a index-antiguide fiber, core index is 1.46, radius is 100, cladding is 1.47, I set the cladding radius 400. The effective index Neff should be complex, but the results are real. The mode field is good.

Could you guys help me fix the problem? Thank you

I attach the .mph file


6 Replies Last Post 16 mar 2012, 10:04 GMT-4
Ivar KJELBERG COMSOL Multiphysics(r) fan, retired, former "Senior Expert" at CSEM SA (CH)

Please login with a confirmed email address before reporting spam

Posted: 1 decade ago 10 gen 2012, 01:15 GMT-5
Hi

are you sure you do not need some dispersion ("k") to get a complex mode ?

--
Good luck
Ivar
Hi are you sure you do not need some dispersion ("k") to get a complex mode ? -- Good luck Ivar

Please login with a confirmed email address before reporting spam

Posted: 1 decade ago 10 gen 2012, 17:19 GMT-5
Hi,

It has no relation to dispersion, just the same as the simple step index fiber (there is a step-index fiber example). The solution I got shows good mode field and correct real part of the effective index, but no imaginary part. You know, this index-antiguided waveguide (core index is smaller than cladding index) is certainly lossy. I used my software to get complex effective index. My boundary just allows no reflection, which can be implemented by PML.

Right now I change to PML, I have a question:
PML is a region we define, then what is the material of the PML, this PML region needs material parameter (permittivity, permeability, elec conductivity). Can I choose air as the PML or I need to set it the same as my adjacent material (say refractive index =1.47 of the cladding), if so I just set permittivity 1.47^2, permeability 1, conductivity 0. Is that right? Or I just use air as the PML layer?

Now another problem is that if I change the thickness of the PML from 50 um -20 um, the imaginary part has a order of 10^3 change which is impossible because 20 um is enough length to absorb. Also if my cladding is very thick, the PML layer does not provide imaginary part (strange). What are the good parameters for the scaling factor and order of PML?

I think the easiest way is you draw a simple index-antiguided fiber model and see the effective index.

Thank you
Hi, It has no relation to dispersion, just the same as the simple step index fiber (there is a step-index fiber example). The solution I got shows good mode field and correct real part of the effective index, but no imaginary part. You know, this index-antiguided waveguide (core index is smaller than cladding index) is certainly lossy. I used my software to get complex effective index. My boundary just allows no reflection, which can be implemented by PML. Right now I change to PML, I have a question: PML is a region we define, then what is the material of the PML, this PML region needs material parameter (permittivity, permeability, elec conductivity). Can I choose air as the PML or I need to set it the same as my adjacent material (say refractive index =1.47 of the cladding), if so I just set permittivity 1.47^2, permeability 1, conductivity 0. Is that right? Or I just use air as the PML layer? Now another problem is that if I change the thickness of the PML from 50 um -20 um, the imaginary part has a order of 10^3 change which is impossible because 20 um is enough length to absorb. Also if my cladding is very thick, the PML layer does not provide imaginary part (strange). What are the good parameters for the scaling factor and order of PML? I think the easiest way is you draw a simple index-antiguided fiber model and see the effective index. Thank you

Ivar KJELBERG COMSOL Multiphysics(r) fan, retired, former "Senior Expert" at CSEM SA (CH)

Please login with a confirmed email address before reporting spam

Posted: 1 decade ago 11 gen 2012, 02:23 GMT-5
Hi

Well still, for me if I take the model library model step index fibre, and look up the equations of the EMW physics, if k=0 hence er=real (and also with sigma=0) how can I get a imaginary eigenmode ?

but if you add the slightest value (>= 1E6*eps) to "k" the modes become immediately complex, or if you add some air around the fiber with some conductivity >0

But OK I might be missing something, and jumping too quickly into the optics, haven't been in this side of COMSOL physics for some time ;)

For the PML, what I remember is that it should be of same material properties as the one on the up/down side of the common boundary, so if you have some air around your glass, then use also air for the PML domain.

Concerning meshing of te PML region, its better to use a structured, layered mesh (you can get that with a circle + layer in V4 geometry, then the geometry will be split up to ease the mapped mesh shape.
for the thickness I use some 10% of total radius for the PML layer, there are several references about adapting the thickness w.r.t. the model in the doc

--
Good luck
Ivar
Hi Well still, for me if I take the model library model step index fibre, and look up the equations of the EMW physics, if k=0 hence er=real (and also with sigma=0) how can I get a imaginary eigenmode ? but if you add the slightest value (>= 1E6*eps) to "k" the modes become immediately complex, or if you add some air around the fiber with some conductivity >0 But OK I might be missing something, and jumping too quickly into the optics, haven't been in this side of COMSOL physics for some time ;) For the PML, what I remember is that it should be of same material properties as the one on the up/down side of the common boundary, so if you have some air around your glass, then use also air for the PML domain. Concerning meshing of te PML region, its better to use a structured, layered mesh (you can get that with a circle + layer in V4 geometry, then the geometry will be split up to ease the mapped mesh shape. for the thickness I use some 10% of total radius for the PML layer, there are several references about adapting the thickness w.r.t. the model in the doc -- Good luck Ivar

Please login with a confirmed email address before reporting spam

Posted: 1 decade ago 11 gen 2012, 17:13 GMT-5
Hi,

I could get imaginary part if my core radius is very small which theoretically gives high loss. But if for large core radius, the comsol could not give imaginary part, maybe FEM is not perfect for this IAG waveguide.

Well, I think right now the problem is basically fixed, but I have another question: do you know how to get fundamental mode quickly, I have too many modes since my structure is large. Is there some sort of auto search fundamental or higher order modes?

Thank you
Hi, I could get imaginary part if my core radius is very small which theoretically gives high loss. But if for large core radius, the comsol could not give imaginary part, maybe FEM is not perfect for this IAG waveguide. Well, I think right now the problem is basically fixed, but I have another question: do you know how to get fundamental mode quickly, I have too many modes since my structure is large. Is there some sort of auto search fundamental or higher order modes? Thank you

Ivar KJELBERG COMSOL Multiphysics(r) fan, retired, former "Senior Expert" at CSEM SA (CH)

Please login with a confirmed email address before reporting spam

Posted: 1 decade ago 11 gen 2012, 18:10 GMT-5
Hi

I do not believe it has to do with the FEM method, rather with the data input and the equations used (Comsol allows to adapt the equations at will, and add specific custom physics).

There is normally a "search close to eignvalue value" to start eigenvalue solving from a given value, check carefully the solver nodes

--
Good luck
Ivar
Hi I do not believe it has to do with the FEM method, rather with the data input and the equations used (Comsol allows to adapt the equations at will, and add specific custom physics). There is normally a "search close to eignvalue value" to start eigenvalue solving from a given value, check carefully the solver nodes -- Good luck Ivar

Please login with a confirmed email address before reporting spam

Posted: 1 decade ago 16 mar 2012, 10:04 GMT-4
hai, i facing some problem related to the complex Neff and the PML.

i wonder how i can get imaginary part for the Neff, since my material 's refractive index is real value, all the inputs parameters are real, how i can get an imaginary value?

then, i try to put PML outside my structure, now, i can get imaginary part for the Neff, may i know why with PML, i can get imaginary part for Neff?

thx
hai, i facing some problem related to the complex Neff and the PML. i wonder how i can get imaginary part for the Neff, since my material 's refractive index is real value, all the inputs parameters are real, how i can get an imaginary value? then, i try to put PML outside my structure, now, i can get imaginary part for the Neff, may i know why with PML, i can get imaginary part for Neff? thx

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.