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.

Insulating spherical ball inside of the constant electric field

Please login with a confirmed email address before reporting spam

I am trying to model a simple textbook problem: insulating spherical ball inside of the constant electric field [Ref: Griffiths Introduction to electrodynamics 3rd ed Chapter 4.4.2]. You can see the example problem from here

From the solution, electric field inside of the insulator is,

Equation shows that the electric field inside of the insulator should be depend on the relative permitivity. However, my simulation shows different result.

I simulate this problem by two stanless steel plate inside of the air. I insert the insulating sphere and change the property of the material. I attached three simulation files; one is with No sphere (the air property of sphere), another one is with the sphere with a different relativive permitivity, and the other one is with the sphere with a different relativie permitivity and conductivity.

From the simulation, I found that the field is depend on the conductivity, but there is no dependence of the relative permivitiy. Could anyone tell me what am I doing wrong?



2 Replies Last Post 7 déc. 2018, 19:29 UTC−5

Please login with a confirmed email address before reporting spam

Posted: 6 years ago 4 déc. 2018, 16:34 UTC−5

I tested different physics [Electrostatics] instead of the [electric current], and it gave me the correct physics which corresponds to the theory. I am still confused why electric current model does not provide accurate physics?

I tested different physics [Electrostatics] instead of the [electric current], and it gave me the correct physics which corresponds to the theory. I am still confused why electric current model does not provide accurate physics?


Robert Koslover Certified Consultant

Please login with a confirmed email address before reporting spam

Posted: 6 years ago 7 déc. 2018, 19:29 UTC−5
Updated: 6 years ago 7 déc. 2018, 19:35 UTC−5

I didn't download your model, but by your own words I can guess what happened. You got the correct answer using the electrostatics model? Good. After all, the problem you were describing is an electrostatics problem, so it is appropriate that the electrostatics module worked, at least if you used it correcly. Now, in regard to the electric current model, if modeling a dc- current case, this is essentially a solution of electric current distributions within media of a finite conductivity. In typical media, this will locally obey a local Ohm's law (J = sigma X E, where J= current density, sigma = conductivity, and E = electric field). As such, the potential differences in your problem (and thus the resulting electric fields in both the media and any additional non-conducting bodies embedded in your media) will depend very much on the resulting current distributions and conductivity properties of that media, since again, your (initial, but mistaken) current model is modeling an Ohm's law problem, not the electrostatic problem that you actually wanted to model. As a result, your partially-conducting media (along with your exterior electrode boundary conditions) imposed its own potentials upon your sphere, and the electric fields within your sphere resulted from that. Varying the dielectric constant of the sphere, but keeping it 100% insulating, evidently did little to impact the current distribution in the media external to your sphere, or the fields and potentials nearby it, per se. But if/when you make it partially conductive, current starts to flow through the sphere too, so now it can actually impact the Ohm's law solution overall. Moral: If you want to study electrostatics in the absence of any currents, don't use a tool intended primarily to solve for currents. After all, you already know/knew that all of those currents (at least, in a correctly-formulated version of the physics problem you mentioned) are/would be equal to zero. :-)

-------------------
Scientific Applications & Research Associates (SARA) Inc.
www.comsol.com/partners-consultants/certified-consultants/sara
I didn't download your model, but by your own words I can guess what happened. You got the correct answer using the electrostatics model? Good. After all, the problem you were describing is an electrostatics problem, so it is appropriate that the electrostatics module worked, at least if you used it correcly. Now, in regard to the electric current model, if modeling a dc- current case, this is essentially a solution of electric current distributions within media of a finite conductivity. In typical media, this will locally obey a local Ohm's law (J = sigma X E, where J= current density, sigma = conductivity, and E = electric field). As such, the potential differences in your problem (and thus the resulting electric fields in both the media and any additional non-conducting bodies embedded in your media) will depend very much on the resulting current distributions and conductivity properties of that media, since again, your (initial, but mistaken) current model is modeling an Ohm's law problem, not the electrostatic problem that you actually wanted to model. As a result, your partially-conducting media (along with your exterior electrode boundary conditions) imposed its own potentials upon your sphere, and the electric fields within your sphere resulted from that. Varying the dielectric constant of the sphere, but keeping it 100% insulating, evidently did little to impact the current distribution in the media external to your sphere, or the fields and potentials nearby it, per se. But if/when you make it partially conductive, current starts to flow through the sphere too, so now it can actually impact the Ohm's law solution overall. Moral: If you want to study electrostatics in the absence of any currents, don't use a tool intended primarily to solve for currents. After all, you already know/knew that all of those currents (at least, in a correctly-formulated version of the physics problem you mentioned) are/would be equal to zero. :-)

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.