Menu

Discussion Forum

New Discussion
Filter

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.

Electric Field Question

Posted Mar 31, 2011, 5:19 a.m. EDT Low-Frequency Electromagnetics, Modeling Tools & Definitions, Parameters, Variables, & Functions 6 Replies

Dominic Clarke
Send Private Message Flag post as spam

Please login with a confirmed email address before reporting spam

Hello,

I have been creating models of electric fields due to objects in air - I usually use a cube or sphere around the object to which i assign the material AIR.

My question is regarding boundary conditions: What is the most physically realistic boundary condition to apply to the outer edge of the AIR in my models? Is is Zero Charge, or Ground or something else?

I have attempted using the infinite elements condition but have always failed there.
6 Replies Last Post Apr 4, 2011, 8:23 a.m. EDT
Louis V
Send Private Message Flag post as spam

Please login with a confirmed email address before reporting spam

Posted: 1 decade ago Mar 31, 2011, 8:27 a.m. EDT
I'm dealing with the same issue, though I just make sure the air large enough so the zero flux boundaries dont effect my electric field too much.
I'm dealing with the same issue, though I just make sure the air large enough so the zero flux boundaries dont effect my electric field too much.
Michael Reilly
Send Private Message Flag post as spam

Please login with a confirmed email address before reporting spam

Posted: 1 decade ago Apr 1, 2011, 12:22 p.m. EDT
In the RF module you should consider using a PML domain surrounding your air domain. On the outside of the PML domain, you can add a scattering boundary condition. This will help to reduce your total domain size for solving in 3d.
In the RF module you should consider using a PML domain surrounding your air domain. On the outside of the PML domain, you can add a scattering boundary condition. This will help to reduce your total domain size for solving in 3d.
Dominic Clarke
Send Private Message Flag post as spam

Please login with a confirmed email address before reporting spam

Posted: 1 decade ago Apr 4, 2011, 5:16 a.m. EDT
Thanks for the responses,

When i use zero charge as a boundary things look completely wrong in the solution - the voltage is spread across the whole space, instead of confined to the metal parts, the electric field doesn't make much sense either.

When I use ground i get a much more well-behaved solution but i wonder if it is right - surrounding my object with ground is like putting it in a faraday cage no? Where no fields can exist.

I would love to have a clear instruction on the best practice here.
Thanks for the responses, When i use zero charge as a boundary things look completely wrong in the solution - the voltage is spread across the whole space, instead of confined to the metal parts, the electric field doesn't make much sense either. When I use ground i get a much more well-behaved solution but i wonder if it is right - surrounding my object with ground is like putting it in a faraday cage no? Where no fields can exist. I would love to have a clear instruction on the best practice here.
Edgar Kaiser
Send Private Message Flag post as spam

Please login with a confirmed email address before reporting spam

Posted: 1 decade ago Apr 4, 2011, 6:11 a.m. EDT
Hi,

as soon as you have electric charge inside the cage, you have fields of course. Only fields from outside sources don't enter the cage.

You will need to make the cage big enough. One way to verify is to check whether the fields within the range you are interested in are still influenced by the size of the cage.
You start with a certain size, run the model, change the size and check the effect on the results. If your desired level of precision is reached you should be fine.

A second option can be to calculate the boundary potential from an analytical approximation. E.g. if your assembly looks like a point charge from the distance you can try to apply the potential distribution of the point charge to the boundary.
This approach works fine for me in a magnetostatic case where I apply the analytical magnetic dipole field to the boundary. This allowed to decrease the size of the container very significantly.

Regards
Edgar
Hi, as soon as you have electric charge inside the cage, you have fields of course. Only fields from outside sources don't enter the cage. You will need to make the cage big enough. One way to verify is to check whether the fields within the range you are interested in are still influenced by the size of the cage. You start with a certain size, run the model, change the size and check the effect on the results. If your desired level of precision is reached you should be fine. A second option can be to calculate the boundary potential from an analytical approximation. E.g. if your assembly looks like a point charge from the distance you can try to apply the potential distribution of the point charge to the boundary. This approach works fine for me in a magnetostatic case where I apply the analytical magnetic dipole field to the boundary. This allowed to decrease the size of the container very significantly. Regards Edgar
Dominic Clarke
Send Private Message Flag post as spam

Please login with a confirmed email address before reporting spam

Posted: 1 decade ago Apr 4, 2011, 7:32 a.m. EDT
Thankyou Edgar,

This is very helpful.

Just to clarify, it is okay for me to use the ground boundary condition (assuming the size of the container no longer affects my solution to with tolerance)

Thanks again,
Kind regards,

Dom.
Thankyou Edgar, This is very helpful. Just to clarify, it is okay for me to use the ground boundary condition (assuming the size of the container no longer affects my solution to with tolerance) Thanks again, Kind regards, Dom.
Edgar Kaiser
Send Private Message Flag post as spam

Please login with a confirmed email address before reporting spam

Posted: 1 decade ago Apr 4, 2011, 8:23 a.m. EDT

It doesn't really matter what you use provided you verify that the boundary doesn't significantly influence your region of interest.
Floating might be better than ground, because it allows some adjustment of the boundary to the potentials the model generates inside.
It doesn't really matter what you use provided you verify that the boundary doesn't significantly influence your region of interest. Floating might be better than ground, because it allows some adjustment of the boundary to the potentials the model generates inside.

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