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.

Non linear conductivity in AC

Please login with a confirmed email address before reporting spam

Hello,
I need to simulate the behavior of a dielectric material in AC to obtain the electric field distribution.
Normally I would use the Electrostatics study, but this material has a nonlinear dependence of the conductivity with the electric field, and the Electrostatics physic doesn't takes in account the influence of conductivity.
I tried to use the Electric currents study with a sinusoidal voltage source and a time dependent solver. The problem is, it takes a lot of time even for a 2D model.
Do you have any ideas on how to improve it? ( or even if my assumptions are wrong, I'm not absolutely sure if EC+time solver is the best model for my problem)

Thank You

3 Replies Last Post 24 mars 2015, 10:08 UTC−4
Edgar J. Kaiser Certified Consultant

Please login with a confirmed email address before reporting spam

Posted: 10 years ago 23 mars 2015, 09:30 UTC−4
Bernardo,

you could try a stationary parametric sweep with the voltage as parameter.

Cheers
Edgar

--
Edgar J. Kaiser
emPhys Physical Technology
www.emphys.com
Bernardo, you could try a stationary parametric sweep with the voltage as parameter. Cheers Edgar -- Edgar J. Kaiser emPhys Physical Technology http://www.emphys.com

Please login with a confirmed email address before reporting spam

Posted: 10 years ago 24 mars 2015, 04:55 UTC−4
Edgar,
Thank you for your answer. The problem is, after I need to take in account the influence of temperature and, for some levels of conductivity, also the charge density. That's why I wanted to go with an temporary solver.

Regards
Bernardo
Edgar, Thank you for your answer. The problem is, after I need to take in account the influence of temperature and, for some levels of conductivity, also the charge density. That's why I wanted to go with an temporary solver. Regards Bernardo

Edgar J. Kaiser Certified Consultant

Please login with a confirmed email address before reporting spam

Posted: 10 years ago 24 mars 2015, 10:08 UTC−4
Bernardo,

ok, then you probably have to keep the time dependent study. You may still try to optimize the solver. Mostly a direct solver is faster than an iterative one, but needs more memory. The nonlinear method under the 'Fully coupled' node can also have a big impact on convergence. Check different nonlinear methods. Check the convergence plot. Sometimes using linear discretization can really speed up a model, but you have to check if accuracy is sufficient. As you are after the electric field, being a spatial derivative, linear elements are probably not a good advice, because the field gets discontinuous. But you can try.

Cheers
Edgar

--
Edgar J. Kaiser
emPhys Physical Technology
www.emphys.com
Bernardo, ok, then you probably have to keep the time dependent study. You may still try to optimize the solver. Mostly a direct solver is faster than an iterative one, but needs more memory. The nonlinear method under the 'Fully coupled' node can also have a big impact on convergence. Check different nonlinear methods. Check the convergence plot. Sometimes using linear discretization can really speed up a model, but you have to check if accuracy is sufficient. As you are after the electric field, being a spatial derivative, linear elements are probably not a good advice, because the field gets discontinuous. But you can try. Cheers Edgar -- Edgar J. Kaiser emPhys Physical Technology http://www.emphys.com

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.