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.

Problem with Mie Model

Please login with a confirmed email address before reporting spam

Hi!
I only got started with COMSOL and was very happy to find the Model 'Mie Scattering off Plasmonic Nanoparticle and Radar Cross-Section Computations'. I followed all the instructions, but when I tried to do the mesh, it wouldn't let me do the swept mesh at first, saying there are involuted elements (even though I followed the instructions to the letter)...
I tried the swept mesh some more and suddenly, it was working fine - until, after 2 hours of calculation time, I saw the results: lots of warnings about an 'inverted mesh element near coordinate (9.24802e-008,9.24802e-008,2.6999e-007)' (these coordinates change slightly when I solve again) and a global variables plot which looks completely unlike the one in the documentation (see attached). What is going on here? I'm at my wit's end...
Thanks in advance for any help!
Kind regards,
Martina


11 Replies Last Post 22 nov. 2011, 12:57 UTC−5

Please login with a confirmed email address before reporting spam

Posted: 1 decade ago 18 janv. 2010, 15:43 UTC−5
Play with meshing more.
Define it manually with care, sometimes automatic meshing give you better results (avoiding inverted elements) but could be not enough for nanoscale.
Play with meshing more. Define it manually with care, sometimes automatic meshing give you better results (avoiding inverted elements) but could be not enough for nanoscale.

Please login with a confirmed email address before reporting spam

Posted: 1 decade ago 15 mars 2010, 02:26 UTC−4
Hi, I am happy to find your example about radar cross section. But I met a problem when I fllowed your example. It happend in Global Expression(dsigma_near, dsigma_far). When I want to plot 4*pi*sigma_back, I met the problem as fllowed: disgma_far is not a valid global experssion. I do not know why.
Hi, I am happy to find your example about radar cross section. But I met a problem when I fllowed your example. It happend in Global Expression(dsigma_near, dsigma_far). When I want to plot 4*pi*sigma_back, I met the problem as fllowed: disgma_far is not a valid global experssion. I do not know why.


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 15 mars 2010, 04:54 UTC−4
Hi

I believe there is a quick reply for your units issue the "r/1[m]" this gives [m] if "r" is without units, or [m^2] as I expect "r" is in [m] and is different from "r/(1[m])" which is without units or equivalent to understanding it as (r[m])/(1[m])

Will have to look closer for the global expression, but I'm a t a Comsol course just now ;) and Ihave to liten to the speaker too ;)

Have a nice day
Ivar
Hi I believe there is a quick reply for your units issue the "r/1[m]" this gives [m] if "r" is without units, or [m^2] as I expect "r" is in [m] and is different from "r/(1[m])" which is without units or equivalent to understanding it as (r[m])/(1[m]) Will have to look closer for the global expression, but I'm a t a Comsol course just now ;) and Ihave to liten to the speaker too ;) Have a nice day Ivar

Please login with a confirmed email address before reporting spam

Posted: 1 decade ago 31 mars 2010, 18:31 UTC−4
Hi to all!
I've got some problems too!
There is serious discrepancy between my results and in results given in the presentation (different field enhancment and cross-sections that are negative at some wavelengths). I used my own data for the refraction index though it also was teken from johnson&christy (1972) paper - I've built interpolated functions with dependence on the wavelength.

Any ideas?

Besides that I can't download the xml library for some reasons (!) and I can't compare dispersion data. Could anybody send it to me. Maybe that will solve the problem...
Thks!


trisha.petrov@gmail.com
Hi to all! I've got some problems too! There is serious discrepancy between my results and in results given in the presentation (different field enhancment and cross-sections that are negative at some wavelengths). I used my own data for the refraction index though it also was teken from johnson&christy (1972) paper - I've built interpolated functions with dependence on the wavelength. Any ideas? Besides that I can't download the xml library for some reasons (!) and I can't compare dispersion data. Could anybody send it to me. Maybe that will solve the problem... Thks! trisha.petrov@gmail.com

Please login with a confirmed email address before reporting spam

Posted: 1 decade ago 16 avr. 2010, 07:13 UTC−4
Hi everyone!

I have the same problem with the model; for some wavelengths the scattering cross section becomes negative which makes, of course, nonsense. First I thought it was probably a mistake I made while defining the global expressions or the integration ones, so I repeated the process several times but I obtained the same result over and over. Later I thought that maybe one of the steps in the documentation file was wrong, so I ended downloading the file .mph directly from the model gallery instead of trying to reproduce it myself. Everything was already defined in the file so I just had to mesh the model and solve it. To do so I followed the instructions (mesh size, number of layers in the exterior subdomain, solver parameters,...) exactly as they were written in the documentation. It was a big surprise to find out that I had exactly the same problem as before. I wanted to attach the scattering cross-sections obtained with the downloaded file (Mie_scattering_plasmonic_reset_35a1.mph) to illustrate the problem but the attach button is not responding. Anyway, any idea about what could be wrong?

Thanks a lot,

Ramon
Hi everyone! I have the same problem with the model; for some wavelengths the scattering cross section becomes negative which makes, of course, nonsense. First I thought it was probably a mistake I made while defining the global expressions or the integration ones, so I repeated the process several times but I obtained the same result over and over. Later I thought that maybe one of the steps in the documentation file was wrong, so I ended downloading the file .mph directly from the model gallery instead of trying to reproduce it myself. Everything was already defined in the file so I just had to mesh the model and solve it. To do so I followed the instructions (mesh size, number of layers in the exterior subdomain, solver parameters,...) exactly as they were written in the documentation. It was a big surprise to find out that I had exactly the same problem as before. I wanted to attach the scattering cross-sections obtained with the downloaded file (Mie_scattering_plasmonic_reset_35a1.mph) to illustrate the problem but the attach button is not responding. Anyway, any idea about what could be wrong? Thanks a lot, Ramon

Please login with a confirmed email address before reporting spam

Posted: 1 decade ago 1 nov. 2010, 10:52 UTC−4
So did the problem of negative cross-section resolve? I am trying to reproduce the results and facing the exact same problem.
So did the problem of negative cross-section resolve? I am trying to reproduce the results and facing the exact same problem.

Please login with a confirmed email address before reporting spam

Posted: 1 decade ago 2 nov. 2010, 06:22 UTC−4
Are you using COMSOL version 3.5?
I had the same problem, but it was solved after installing the latest hotfix.
Are you using COMSOL version 3.5? I had the same problem, but it was solved after installing the latest hotfix.

Please login with a confirmed email address before reporting spam

Posted: 1 decade ago 25 août 2011, 03:37 UTC−4
Hi, every one. Has anyone realized the model in v4.2?

As the variables in v4.2 is different with v3.5. I can not perform the simulation in v4.2.

I found on variables named as scEx, scHy, etc.

It made me rather confusing.

Can anyone help me?
Hi, every one. Has anyone realized the model in v4.2? As the variables in v4.2 is different with v3.5. I can not perform the simulation in v4.2. I found on variables named as scEx, scHy, etc. It made me rather confusing. Can anyone help me?

Please login with a confirmed email address before reporting spam

Posted: 1 decade ago 22 nov. 2011, 11:07 UTC−5
Hi,

I am trying to use the variables defined in the Mie scattering tutorial and have found out that sc should be replaced with mod1.rfw.rel, for the E components and just rfw. for the H components. For example 0.5*real(mod1.rfw.relEx*0.25*conj(relHz)-mod1.rfw.relEy*0.25*conj(relHx)). The power variables are called Poavx instead of Poxav and the sc should again be replaced with rel. Poxav_rfw is mod1.rfw.Poavx in v4.2a.

But even though I have changed this in my model I can't get it to work because I still get the error "unexpected unit of input" for the dsigma_far variable : (normEfar^2/(r/(1[m]))^2)/abs(E0)^2*r^2 where r is defined as sqrt(x^2+y^2+z^2)

In equation view under far field calculation tab I see that the normEfar is in V/m units and so is E0 so the expression should give me units of m^2 which is what I want.

Does anybody know how to fix this error?

thanks in advance,
Signe
Hi, I am trying to use the variables defined in the Mie scattering tutorial and have found out that sc should be replaced with mod1.rfw.rel, for the E components and just rfw. for the H components. For example 0.5*real(mod1.rfw.relEx*0.25*conj(relHz)-mod1.rfw.relEy*0.25*conj(relHx)). The power variables are called Poavx instead of Poxav and the sc should again be replaced with rel. Poxav_rfw is mod1.rfw.Poavx in v4.2a. But even though I have changed this in my model I can't get it to work because I still get the error "unexpected unit of input" for the dsigma_far variable : (normEfar^2/(r/(1[m]))^2)/abs(E0)^2*r^2 where r is defined as sqrt(x^2+y^2+z^2) In equation view under far field calculation tab I see that the normEfar is in V/m units and so is E0 so the expression should give me units of m^2 which is what I want. Does anybody know how to fix this error? thanks in advance, Signe

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 22 nov. 2011, 11:18 UTC−5
Hi
Indeed naming has changed quite a lot between 3.5 and v4.
For the units, do not forget that orange unit flagging is only a warning, it will solve in any case, (but it might be wrong)

it could be that abs() is not passing the units, try it out in a post processing field

--
Good luck
Ivar
Hi Indeed naming has changed quite a lot between 3.5 and v4. For the units, do not forget that orange unit flagging is only a warning, it will solve in any case, (but it might be wrong) it could be that abs() is not passing the units, try it out in a post processing field -- Good luck Ivar

Please login with a confirmed email address before reporting spam

Posted: 1 decade ago 22 nov. 2011, 12:57 UTC−5
Hi,

I tried to run my program but I don't get any data for the different cross sections. I'm not trying to replicate the Mie scattering tutorial but rather trying to use some of the variables in my own model with a gold nano cavity so I can get the spectra. I think there is a problem with my integration and the parametric sweep over wavelengths, is it possible you can take a quick look at it?
Do you know what the best way is to obtain a reflectivity spectrum?

regards Signe
Hi, I tried to run my program but I don't get any data for the different cross sections. I'm not trying to replicate the Mie scattering tutorial but rather trying to use some of the variables in my own model with a gold nano cavity so I can get the spectra. I think there is a problem with my integration and the parametric sweep over wavelengths, is it possible you can take a quick look at it? Do you know what the best way is to obtain a reflectivity spectrum? regards Signe

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.