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.

negative scattering cross sections

Please login with a confirmed email address before reporting spam

Hi,

This might be a silly and old question: I got negative scattering cross sections in a simulation of gold nanorod. The light source is linearly polarized plane wave, the puzzling part is that if the incidnet electric field is along the long axis, I got expected optical response (absorption and scattering cross sections), however, if I change the polarization angle to perpendicular to the long axis, the absorption looks fine but the scattering cross section becomes negative.

The scattering crosssection is calculated as:

  1. create a variable, let's call it A, which expression is nx * ewfd.relPoavx + ny * ewfd.relPoavy + nz * ewfd.relPoavz
  2. integrate over the boudaries of the nanorod and then divided by the incident power density P0 (unit: W/m^2) sigma_scat = intep_Surf(A) / P0

Any help or discussion is much appreciated!


3 Replies Last Post 31 juil. 2023, 17:57 UTC−4
Robert Koslover Certified Consultant

Please login with a confirmed email address before reporting spam

Posted: 1 year ago 23 juil. 2023, 18:38 UTC−4
Updated: 1 year ago 23 juil. 2023, 19:07 UTC−4
  1. Wires perpendicular to an incident polarization usually won't scatter much. Is your negative cross section merely numerical noise? How does its magnitude compare to the cross section for parallel incident polarization?
  2. Try remeshing the scatterer more finely and see if you get the same result.
  3. Recheck all your boundary conditions and material properties and definitions, etc.
  4. Make sure your nx, ny, and nz unit vectors all point in the correct directions at the interfaces on which they are being used (do a surface arrow plot of nx, ny, nz on those surfaces).
  5. You might want to consider posting your .mph file to the forum.
-------------------
Scientific Applications & Research Associates (SARA) Inc.
www.comsol.com/partners-consultants/certified-consultants/sara
1. Wires perpendicular to an incident polarization usually won't scatter much. Is your negative cross section merely numerical noise? How does its magnitude compare to the cross section for parallel incident polarization? 2. Try remeshing the scatterer more finely and see if you get the same result. 3. Recheck all your boundary conditions and material properties and definitions, etc. 4. Make sure your nx, ny, and nz unit vectors all point in the correct directions at the interfaces on which they are being used (do a surface arrow plot of nx, ny, nz on those surfaces). 5. You might want to consider posting your .mph file to the forum.

Please login with a confirmed email address before reporting spam

Posted: 1 year ago 31 juil. 2023, 11:33 UTC−4

Thank you Robert. I will go through your solutions and I will post my .mph file if I need further help. Appreciate your comments, it is very helpful.

Thank you Robert. I will go through your solutions and I will post my .mph file if I need further help. Appreciate your comments, it is very helpful.

Please login with a confirmed email address before reporting spam

Posted: 1 year ago 31 juil. 2023, 17:57 UTC−4

Hi, Robert and other folks.

  1. I checked the result, it is definitely not numerical noise, it looks like expected spectrum with reversed sign.

  2. I've try finer mesh but still got negative cross sections.

  3. I assume that my boundary conditions and material properties are fine, because it works well for parallel polarization.

  4. I plot surface arrow plot of the nx, ny, nz and the ewfd.relPoavx, etc, they seems both point outward of the boundary, so mathematically I should get positive result.

  5. The .mph file is shared. Please take a look and I greatly appreciate your time and effort.

Hi, Robert and other folks. 1. I checked the result, it is definitely not numerical noise, it looks like expected spectrum with reversed sign. 2. I've try finer mesh but still got negative cross sections. 3. I assume that my boundary conditions and material properties are fine, because it works well for parallel polarization. 4. I plot surface arrow plot of the nx, ny, nz and the ewfd.relPoavx, etc, they seems both point outward of the boundary, so mathematically I should get positive result. 5. The .mph file is shared. Please take a look and I greatly appreciate your time and effort.

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.