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.

Complex conductivity in RF transient model and absorption boundaries

Please login with a confirmed email address before reporting spam

I have been studying a Gaussian beam incidents on an absorbing material. The result is reasonable before the wave hits the material, but starting to give strange results afterwards. Especially, when the simulation time is slightly longer, the field begins to increase dramatically. I only input an electric field of 10^5V/m, but the results indicates an electric field of 10^56V/m all over the model at the end.

I guess the problem might come from the following two aspects.

In transient solver, there is only one blank to define the conductivity. For absorbing material, the conductivity is complex. Do anyone know if it is OK to put a complex number in the blank?

Ideally the boundary should be perfectly absorbing with no reflection. I use scattering boundary conditions because there is no option for PML as far as I can see. Maybe some resonance occurs which leads to the problem? Is there anyway to build a perfectly absorbing boundary in transient RF solver?

3 Replies Last Post 2016年8月31日 GMT-4 14:43
Robert Koslover Certified Consultant

Please login with a confirmed email address before reporting spam

Posted: 9 years ago 2015年11月21日 GMT-5 10:48
A comment on the real vs. complex-value part of your question: I may be mistaken, but from a purely physics perspective, when working in the time domain, I believe one would never use a complex conductivity. The use of complex numbers in classical EM arises due to the frequency-domain formulation and the convenience of replacing sines and cosines with exponentials. This does not apply in a purely time-domain formulation. In time domain, all physical quantities and properties should be represented by real values.
A comment on the real vs. complex-value part of your question: I may be mistaken, but from a purely physics perspective, when working in the time domain, I believe one would never use a complex conductivity. The use of complex numbers in classical EM arises due to the frequency-domain formulation and the convenience of replacing sines and cosines with exponentials. This does not apply in a purely time-domain formulation. In time domain, all physical quantities and properties should be represented by real values.

Please login with a confirmed email address before reporting spam

Posted: 9 years ago 2015年11月23日 GMT-5 07:26
Hi,

Thanks for your advice. I think what you said make sense in some extent. To my understanding, the imaginary part of the conductivity doesn't stand for the loss but just a phase difference between the driving field and the current, so it is safe to remove the imaginary part the conductivity in my case.

Still, even if I define the conductivity as real parts only, the wired strong electric field remains to appear. The only way to avoid this I found up to now is to remove the imaginary part of the permittivity (refractive index). This is obviously not acceptable as I am simulating transient respond of absorbing materials.

Do you have any experience and suggestions regarding on this? I am totally stuck at here.
Hi, Thanks for your advice. I think what you said make sense in some extent. To my understanding, the imaginary part of the conductivity doesn't stand for the loss but just a phase difference between the driving field and the current, so it is safe to remove the imaginary part the conductivity in my case. Still, even if I define the conductivity as real parts only, the wired strong electric field remains to appear. The only way to avoid this I found up to now is to remove the imaginary part of the permittivity (refractive index). This is obviously not acceptable as I am simulating transient respond of absorbing materials. Do you have any experience and suggestions regarding on this? I am totally stuck at here.

Please login with a confirmed email address before reporting spam

Posted: 8 years ago 2016年8月31日 GMT-4 14:43
Did you find any ways after that how to use complex number in time domain ? I have similar issues here as well. I need to define complex permittivity material which interacts with plane waves in time domain.

Please let me know if you have any suggestion Robert and Zeng.

Thanks,
RA
Did you find any ways after that how to use complex number in time domain ? I have similar issues here as well. I need to define complex permittivity material which interacts with plane waves in time domain. Please let me know if you have any suggestion Robert and Zeng. Thanks, RA

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.