Menu

Discussion Forum

New Discussion
Filtern

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 The information provided may be out of date.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.

PCF design

Posted 09.01.2012, 06:21 GMT-5 RF & Microwave Engineering, Wave Optics Version 4.4 30 Replies

Selim Habib
Send Private Message Flag post as spam

Please login with a confirmed email address before reporting spam

How can I get imaginary part of effective mode index using comsol4.0 ? Help me.
30 Replies Last Post 22.03.2017, 07:39 GMT-4
Ivar KJELBERG COMSOL Multiphysics(r) fan, retired, former "Senior Expert" at CSEM SA (CH)
Send Private Message Flag post as spam

Please login with a confirmed email address before reporting spam

Posted: 1 decade ago 09.01.2012, 06:41 GMT-5
Hi

ave you tried a search for "imaginary part mode" on the Forum and the model library ?.
Note, you need to have some damping/loss properties to get an imaginary part

--
Good luck
Ivar
Hi ave you tried a search for "imaginary part mode" on the Forum and the model library ?. Note, you need to have some damping/loss properties to get an imaginary part -- Good luck Ivar
Selim Habib
Send Private Message Flag post as spam

Please login with a confirmed email address before reporting spam

Posted: 1 decade ago 09.01.2012, 07:21 GMT-5
Hi Ivar,

Thanks for your immediate reply.

Actually i got real part of the refractive index and using 5 ring airholes (hexagonal pattern).
Would you please more clarify me to get the imaginary part of mode index.


Thanx.
Hi Ivar, Thanks for your immediate reply. Actually i got real part of the refractive index and using 5 ring airholes (hexagonal pattern). Would you please more clarify me to get the imaginary part of mode index. Thanx.
Ivar KJELBERG COMSOL Multiphysics(r) fan, retired, former "Senior Expert" at CSEM SA (CH)
Send Private Message Flag post as spam

Please login with a confirmed email address before reporting spam

Posted: 1 decade ago 09.01.2012, 08:22 GMT-5
Hi

from my understanding you need an complex index (n,k) or some "k" different from "0" to get a complex mode, no ?
I'm not yet familiar with the new feature "Solver Split complex & real parts" but if you leave it unchecked, and add a little "k" to your material, your effective mode index becomes immediately complex too, than you can take the real and imaginary part thereof. I.e. try it out on the model library example

--
Good luck
Ivar
Hi from my understanding you need an complex index (n,k) or some "k" different from "0" to get a complex mode, no ? I'm not yet familiar with the new feature "Solver Split complex & real parts" but if you leave it unchecked, and add a little "k" to your material, your effective mode index becomes immediately complex too, than you can take the real and imaginary part thereof. I.e. try it out on the model library example -- Good luck Ivar
Selim Habib
Send Private Message Flag post as spam

Please login with a confirmed email address before reporting spam

Posted: 1 decade ago 09.01.2012, 23:24 GMT-5
Thanx Ivar.
Thanx Ivar.
Selim Habib
Send Private Message Flag post as spam

Please login with a confirmed email address before reporting spam

Posted: 1 decade ago 10.01.2012, 10:28 GMT-5
Hi

I used core as silica glass doped and air holes. I set real part of ref. index of silica glass doped as 1.4457 and air as 1 and used four hexagonal pattern. Good mode field occurs and got corresponding real part of effective mode index. Which initial value of imaginary part of ref. index(silica glass doped and air) i will set up to get imaginary part of effective mode index.
Hi I used core as silica glass doped and air holes. I set real part of ref. index of silica glass doped as 1.4457 and air as 1 and used four hexagonal pattern. Good mode field occurs and got corresponding real part of effective mode index. Which initial value of imaginary part of ref. index(silica glass doped and air) i will set up to get imaginary part of effective mode index.
Ivar KJELBERG COMSOL Multiphysics(r) fan, retired, former "Senior Expert" at CSEM SA (CH)
Send Private Message Flag post as spam

Please login with a confirmed email address before reporting spam

Posted: 1 decade ago 10.01.2012, 11:12 GMT-5
Hi

you need the full index of refraction "(n+j*k) or the complex dielectric constant "er" (not the chromatic dispersion of
"n(lambda)"
If I remember right "k" very low <1E-4 in the visible window check your optical handbooks ;)

--
Good luck
Ivar
Hi you need the full index of refraction "(n+j*k) or the complex dielectric constant "er" (not the chromatic dispersion of "n(lambda)" If I remember right "k" very low
Selim Habib
Send Private Message Flag post as spam

Please login with a confirmed email address before reporting spam

Posted: 1 decade ago 23.01.2012, 02:20 GMT-5
Hi,
I got the real part of effective mode index 1.425868(say). that means six digits after decimal points. how can i get more digits after decimal points. help me. i used comsol 4 version.
Hi, I got the real part of effective mode index 1.425868(say). that means six digits after decimal points. how can i get more digits after decimal points. help me. i used comsol 4 version.
Ivar KJELBERG COMSOL Multiphysics(r) fan, retired, former "Senior Expert" at CSEM SA (CH)
Send Private Message Flag post as spam

Please login with a confirmed email address before reporting spam

Posted: 1 decade ago 23.01.2012, 04:14 GMT-5
Hi

if its in a table you have an icon switch option that gives you "full" precision

But that does not mean the other digits are correct, you need to check the sensitivity of your model to mesh density, solver settings (particularly absolute and relative tolerances), material parameters geometry dimensions to really decide how many digits are "correct". And finally what is the numerical precision you can expect from double precision 64 bit binary numbers solving 2nd order equations (but that's really when you arrive from the last 5-6 digits to the right

--
Good luck
Ivar
Hi if its in a table you have an icon switch option that gives you "full" precision But that does not mean the other digits are correct, you need to check the sensitivity of your model to mesh density, solver settings (particularly absolute and relative tolerances), material parameters geometry dimensions to really decide how many digits are "correct". And finally what is the numerical precision you can expect from double precision 64 bit binary numbers solving 2nd order equations (but that's really when you arrive from the last 5-6 digits to the right -- Good luck Ivar
Selim Habib
Send Private Message Flag post as spam

Please login with a confirmed email address before reporting spam

Posted: 1 decade ago 12.02.2012, 01:04 GMT-5
Hi Ivar,
I want to get birefringence B= nx-ny. How can i get nx and ny in comsol 4.2 version.
Hi Ivar, I want to get birefringence B= nx-ny. How can i get nx and ny in comsol 4.2 version.
Ivar KJELBERG COMSOL Multiphysics(r) fan, retired, former "Senior Expert" at CSEM SA (CH)
Send Private Message Flag post as spam

Please login with a confirmed email address before reporting spam

Posted: 1 decade ago 12.02.2012, 06:28 GMT-5
Hi

interesting question, havent tried, but there shuld be a linear or anistropic material feature, at least in the epsilonr form

--
Good luck
Ivar
Hi interesting question, havent tried, but there shuld be a linear or anistropic material feature, at least in the epsilonr form -- Good luck Ivar
Selim Habib
Send Private Message Flag post as spam

Please login with a confirmed email address before reporting spam

Posted: 1 decade ago 14.02.2012, 00:54 GMT-5
Thanks for the reply.
Thanks for the reply.
SIVACOUMAR RAJALINGAM
Send Private Message Flag post as spam

Please login with a confirmed email address before reporting spam

Posted: 1 decade ago 23.02.2012, 11:07 GMT-5
In the Postprocessing after getting the confinement, go to general and see for the effective mode index. There will be effective mode index with maximum confinement, which can be confirmed by the imaginary part of the mode index. Two effective mode index with maximum confinement is taken as nx and ny . To confirm that use arrow plot for those two mode index. You can see the arrow plot will be in different direction ie. 90. Hope it will help u.
In the Postprocessing after getting the confinement, go to general and see for the effective mode index. There will be effective mode index with maximum confinement, which can be confirmed by the imaginary part of the mode index. Two effective mode index with maximum confinement is taken as nx and ny . To confirm that use arrow plot for those two mode index. You can see the arrow plot will be in different direction ie. 90. Hope it will help u.

Attachments:

Selim Habib
Send Private Message Flag post as spam

Please login with a confirmed email address before reporting spam

Posted: 1 decade ago 27.02.2012, 10:32 GMT-5
Hi,
Thank you very much for you reply.
I took the values of neff for the maximum confinement.
I got the values as: neff(1)= 1.234501473939883 and neff(2)=1.2345017246688192
and the arrow plot is 90 degree as you said.


Please confirm me about my information.
Hi, Thank you very much for you reply. I took the values of neff for the maximum confinement. I got the values as: neff(1)= 1.234501473939883 and neff(2)=1.2345017246688192 and the arrow plot is 90 degree as you said. Please confirm me about my information.
Tee Din Chai
Send Private Message Flag post as spam

Please login with a confirmed email address before reporting spam

Posted: 1 decade ago 16.03.2012, 10:15 GMT-4
hai, so, finally, are you able to find out the imaginary part of Neff?

if you solved it, may i know how you solve it?
because i am facing the same problems.

thx
hai, so, finally, are you able to find out the imaginary part of Neff? if you solved it, may i know how you solve it? because i am facing the same problems. thx
Selim Habib
Send Private Message Flag post as spam

Please login with a confirmed email address before reporting spam

Posted: 1 decade ago 18.03.2012, 02:45 GMT-4
Hi,
You have to set PML to get the imaginary part.
Hi, You have to set PML to get the imaginary part.
Sathiyan S
Send Private Message Flag post as spam

Please login with a confirmed email address before reporting spam

Posted: 1 decade ago 26.05.2014, 13:14 GMT-4
Hi Selim Habib ,
I have one doubt.How did u get 16 decimal value of effective index value ?. Because in COMSOL we can able to get only 6 decimal values know ? Please help me in this issue. Many thanks in advance !




Thanks
Hi Selim Habib , I have one doubt.How did u get 16 decimal value of effective index value ?. Because in COMSOL we can able to get only 6 decimal values know ? Please help me in this issue. Many thanks in advance ! Thanks
Md Selim Habib
Send Private Message Flag post as spam

Please login with a confirmed email address before reporting spam

Posted: 1 decade ago 26.05.2014, 13:23 GMT-4
Hi,

Use full precision.
Hi, Use full precision.
Ryan Wang
Send Private Message Flag post as spam

Please login with a confirmed email address before reporting spam

Posted: 10 years ago 24.09.2014, 12:22 GMT-4
Hey, excuse me but could you tell me the detailed process of achieving birefringence of a known pcf design?

I can just get effective refractive index using mode analysis under the Wave Optics module but I don't konw if that is fundamental mode or not and how can I get nx and ny ?

And I saw you said using PML produces the imaginary part of refractive index which can help confirming fundamental mode, how should I set it?

Looking forwards to your reply, thanks a lot!
Hey, excuse me but could you tell me the detailed process of achieving birefringence of a known pcf design? I can just get effective refractive index using mode analysis under the Wave Optics module but I don't konw if that is fundamental mode or not and how can I get nx and ny ? And I saw you said using PML produces the imaginary part of refractive index which can help confirming fundamental mode, how should I set it? Looking forwards to your reply, thanks a lot!
Md Selim Habib
Send Private Message Flag post as spam

Please login with a confirmed email address before reporting spam

Posted: 10 years ago 24.09.2014, 13:57 GMT-4
Hi Wang,

Could you please send me your model?
Hi Wang, Could you please send me your model?
Jia Mao
Send Private Message Flag post as spam

Please login with a confirmed email address before reporting spam

Posted: 9 years ago 17.03.2015, 02:14 GMT-4
hi
have you solve your problem about imaginary part, I have same problem.how to??
hi have you solve your problem about imaginary part, I have same problem.how to??
Limin Wu
Send Private Message Flag post as spam

Please login with a confirmed email address before reporting spam

Posted: 9 years ago 28.06.2015, 23:14 GMT-4
hi
have you solve your problem about imaginary part, I have same problem.how to??
I cannt get the complex index of the fundmental mode even after the PML is applied. It is always a real number.
hi have you solve your problem about imaginary part, I have same problem.how to?? I cannt get the complex index of the fundmental mode even after the PML is applied. It is always a real number.
Limin Wu
Send Private Message Flag post as spam

Please login with a confirmed email address before reporting spam

Posted: 9 years ago 28.06.2015, 23:31 GMT-4
hi, Selim Habib
have you solve the problem about imaginary part? I have same problem. I cannt get the complex index of the fundmental mode even after the PML is applied. It is always a real number.
hi, Selim Habib have you solve the problem about imaginary part? I have same problem. I cannt get the complex index of the fundmental mode even after the PML is applied. It is always a real number.

Attachments:

Nafiz
Send Private Message Flag post as spam

Please login with a confirmed email address before reporting spam

Posted: 9 years ago 20.07.2015, 02:18 GMT-4
Hi

Did you get around the problem of getting zero in the imaginary part of effective index ? Was it by applying a small imaginary refractive index for silica ? what was the value ?
Hi Did you get around the problem of getting zero in the imaginary part of effective index ? Was it by applying a small imaginary refractive index for silica ? what was the value ?
Jiaqi Ye
Send Private Message Flag post as spam

Please login with a confirmed email address before reporting spam

Posted: 9 years ago 01.01.2016, 08:29 GMT-5
Hi, did you solve the problem of getting zero in the imaginary part of effective index? I suffer from this problem a lot, it would be nice if you can help. Thanks.
Hi, did you solve the problem of getting zero in the imaginary part of effective index? I suffer from this problem a lot, it would be nice if you can help. Thanks.
Tee Din Chai
Send Private Message Flag post as spam

Please login with a confirmed email address before reporting spam

Posted: 9 years ago 01.01.2016, 08:36 GMT-5
To get imaginary part of Neff, you need to apply Perfectly matched layer to truncate your structure.
I feel that the default PML in comsol is not accurate. Might be my mistake. You can try anisotropy PML.

alternative, if your material refractive index is with imaginary part for example 1.45+1i, then, without using PML, you will get Imaginary Neff.
If your material refractive index is purely real, then, you will forever won't get Imaginary Neff if you didn't use PML in your mode analysis.

Hope this helps.
To get imaginary part of Neff, you need to apply Perfectly matched layer to truncate your structure. I feel that the default PML in comsol is not accurate. Might be my mistake. You can try anisotropy PML. alternative, if your material refractive index is with imaginary part for example 1.45+1i, then, without using PML, you will get Imaginary Neff. If your material refractive index is purely real, then, you will forever won't get Imaginary Neff if you didn't use PML in your mode analysis. Hope this helps.
Jiaqi Ye
Send Private Message Flag post as spam

Please login with a confirmed email address before reporting spam

Posted: 9 years ago 01.01.2016, 08:47 GMT-5
Thanks for your immediate reply. The thing is I have setup the PML layer and get the imaginary part for most of the effective indexes. But the problem is, for the optimal effective index with minimum value of imaginary part, I can not see the exact value of imaginary part. I guess it may be a matter of the value is too small, e.g. <E-12i, so it is not displayed. The attach file is a screenshot of the effective index I got.

In the Comsol3.5 version, we can get the exact value via generate the report, but I don't know how to deal with it in COMSOL4.4. Any experience with this?
Thanks for your immediate reply. The thing is I have setup the PML layer and get the imaginary part for most of the effective indexes. But the problem is, for the optimal effective index with minimum value of imaginary part, I can not see the exact value of imaginary part. I guess it may be a matter of the value is too small, e.g.

Attachments:

Tee Din Chai
Send Private Message Flag post as spam

Please login with a confirmed email address before reporting spam

Posted: 9 years ago 01.01.2016, 08:56 GMT-5
This is because the Imag(Neff) is too small. Loss is negligible.
You can see the full precision at comsol UI. On the bottom right (the table..), you generate your Neff and click the full precision.

Or if you have comsol matlab, you can see the full precision also.
This is because the Imag(Neff) is too small. Loss is negligible. You can see the full precision at comsol UI. On the bottom right (the table..), you generate your Neff and click the full precision. Or if you have comsol matlab, you can see the full precision also.
Jiaqi Ye
Send Private Message Flag post as spam

Please login with a confirmed email address before reporting spam

Posted: 9 years ago 01.01.2016, 09:25 GMT-5
I have tried evaluate neff in the table with full precious, but it just display the full precious of the real part, for the imaginary part, it is still not viable. See the attache file for the detail.
I have tried evaluate neff in the table with full precious, but it just display the full precious of the real part, for the imaginary part, it is still not viable. See the attache file for the detail.

Attachments:

Tee Din Chai
Send Private Message Flag post as spam

Please login with a confirmed email address before reporting spam

Posted: 9 years ago 01.01.2016, 21:29 GMT-5
get the imaginary part of neff.
imag(emw.neff)

i prefer to use comsol matlab, where you can extract the data directly.
get the imaginary part of neff. imag(emw.neff) i prefer to use comsol matlab, where you can extract the data directly.
Mahmood Waqas
Send Private Message Flag post as spam

Please login with a confirmed email address before reporting spam

Posted: 7 years ago 22.03.2017, 07:39 GMT-4
Dear Selim,

Which components of Electric Field did you solve for? Three components or in-plane vector. Please leave a reply.
Regards,
Waqas.
Dear Selim, Which components of Electric Field did you solve for? Three components or in-plane vector. Please leave a reply. Regards, Waqas.

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.

Suggested Content