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.

Electrostatic model

Please login with a confirmed email address before reporting spam

Hello all,

Hopefully this is a simple question. I'm trying to calculate some forces in the electrostatic model and I've been running into trouble. This is a rotationally symmetric model consisting of an electrode at a fixed voltage in vacuum above a dielectric material over a ground plane.

To debug the problems that I've been having, I've tried to simply perform a surface integral over the electrode of the surface charge.

Value of surface integral: 1.694352e-18 [C], Expression: -(Dz_emes*nz + Dr_emes*nr), Boundary: 8
Value of surface integral: 3.388705e-18 [C], Expression: nD_emes, Boundary: 8

these are the results. According to documentation, the surface charge nD_emes is calculated by finding -n dot D, which is what is on the first line above.

I'm wondering why the two results are not the same?

Thanks,
Everet

6 Replies Last Post 4 déc. 2010, 07:33 UTC−5
Ivar KJELBERG COMSOL Multiphysics(r) fan, retired, former "Senior Expert" at CSEM SA (CH)

Please login with a confirmed email address before reporting spam

Posted: 1 decade ago 1 déc. 2010, 14:29 UTC−5
Hi

you have a very good question there,

but as the difference looks like a factor "2" could it be a rms value or simply a true bug, there were a few typos in the equations in ACDC for 4.0a, corrected now in 4.1

Take a look at the equtions beneath and check if you find something wrong, check the model in 4.1 (if you are not already with the latest patch of 4.1) if still there report it to "support" they shpould be able to explain

--
Good luck
Ivar
Hi you have a very good question there, but as the difference looks like a factor "2" could it be a rms value or simply a true bug, there were a few typos in the equations in ACDC for 4.0a, corrected now in 4.1 Take a look at the equtions beneath and check if you find something wrong, check the model in 4.1 (if you are not already with the latest patch of 4.1) if still there report it to "support" they shpould be able to explain -- Good luck Ivar

Please login with a confirmed email address before reporting spam

Posted: 1 decade ago 1 déc. 2010, 15:34 UTC−5
Actually the particular machine I'm using is running v3.5. I think there are some computers with a different version - I'll check them out.

Thanks for your input
Actually the particular machine I'm using is running v3.5. I think there are some computers with a different version - I'll check them out. Thanks for your input

Ivar KJELBERG COMSOL Multiphysics(r) fan, retired, former "Senior Expert" at CSEM SA (CH)

Please login with a confirmed email address before reporting spam

Posted: 1 decade ago 1 déc. 2010, 15:58 UTC−5
Hi

you can still look at the equations (in 3,5), you need to find the correct path via the windows and pull down menus.

But I cannot remember of any equation typos in 3.5 so I suspect you are doing something slightly different for the two cases, and you are missing a 0.5 or a 2 somewhere

--
Good luck
Ivar
Hi you can still look at the equations (in 3,5), you need to find the correct path via the windows and pull down menus. But I cannot remember of any equation typos in 3.5 so I suspect you are doing something slightly different for the two cases, and you are missing a 0.5 or a 2 somewhere -- Good luck Ivar

Please login with a confirmed email address before reporting spam

Posted: 1 decade ago 3 déc. 2010, 04:54 UTC−5
Hi,
Interesting but
I have tried the same thing as you (in my own model): These are my results

Value of surface integral: -3.613317e-17 [C], Expression: -(Dz_emes*nz + Dr_emes*nr), Boundary: 2
Value of surface integral: -3.613317e-17 [C], Expression: nD_emes, Boundary: 2

Check your model again!!
Good luck

Hi, Interesting but I have tried the same thing as you (in my own model): These are my results Value of surface integral: -3.613317e-17 [C], Expression: -(Dz_emes*nz + Dr_emes*nr), Boundary: 2 Value of surface integral: -3.613317e-17 [C], Expression: nD_emes, Boundary: 2 Check your model again!! Good luck

Please login with a confirmed email address before reporting spam

Posted: 1 decade ago 3 déc. 2010, 11:59 UTC−5
Thanks for your thoughts.

I just tried this again, with different (but inconsistent) results. I built the model again, meshed and solved it then performed the surface integral over the electrode.

Value of surface integral: 1.045e-18 [C], Expression: -(Dz_emes*nz + Dr_emes*nr), Boundaries: 8, 10, 12
Value of surface integral: 8.467133e-18 [C], Expression: nD_emes, Boundaries: 8, 10, 12

I looked at the equations and found that nD_emes = unr*(down(Dr_emes)-up(Dr_emes))+unz*(down(Dz_emes)-up(Dz_emes))

To be perfectly honest, I don't see how nD_emes = -del dot D. To be fair though, I don't think I fully understand what the up() and down() functions are for.

Insight may be all I need,
Thanks for all the help,
Everet
Thanks for your thoughts. I just tried this again, with different (but inconsistent) results. I built the model again, meshed and solved it then performed the surface integral over the electrode. Value of surface integral: 1.045e-18 [C], Expression: -(Dz_emes*nz + Dr_emes*nr), Boundaries: 8, 10, 12 Value of surface integral: 8.467133e-18 [C], Expression: nD_emes, Boundaries: 8, 10, 12 I looked at the equations and found that nD_emes = unr*(down(Dr_emes)-up(Dr_emes))+unz*(down(Dz_emes)-up(Dz_emes)) To be perfectly honest, I don't see how nD_emes = -del dot D. To be fair though, I don't think I fully understand what the up() and down() functions are for. Insight may be all I need, Thanks for all the help, Everet

Ivar KJELBERG COMSOL Multiphysics(r) fan, retired, former "Senior Expert" at CSEM SA (CH)

Please login with a confirmed email address before reporting spam

Posted: 1 decade ago 4 déc. 2010, 07:33 UTC−5
Hi

up and down are the normal of a boundary looking into one and the other domains. If you have physicsl flux continuity at the boundary they should give you similar results (up to a +/- sign definition) but if you have a discontinuity at the boundary your results would only be correct if you consder the differences of each domain, that is what the specific "up" and "down" are for.

plot them out for different cases (when they exist) and you will rapidly catch their interest

--
Good luck
Ivar
Hi up and down are the normal of a boundary looking into one and the other domains. If you have physicsl flux continuity at the boundary they should give you similar results (up to a +/- sign definition) but if you have a discontinuity at the boundary your results would only be correct if you consder the differences of each domain, that is what the specific "up" and "down" are for. plot them out for different cases (when they exist) and you will rapidly catch their interest -- Good luck Ivar

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.