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.

What about Maxwell's equation ?

Please login with a confirmed email address before reporting spam

Hi everybody,

In order to solve my problem (see : www.comsol.fr/community/forums/general/thread/6526/), i have tried a new approach. COMSOL seems to need some potential as reference to find solutions in electrostatics. As i'm interested in electric fields and not potential, i have implemented the maxwell equation (Div E = rho /eps_0) in my system using
PDE in coefficient. (parameters c=a=e=d=alpha=gamma=0 ; beta=1 ; f=0 or XXX). My charge pattern could be a source or have a "source" boundary. But when i tried to plot the resulted electric field E, COMSOL can't find a solution...
Is the way i solve/implement it right ? Do you have any suggestion ?

Warm thanks for your help.

Etienne

3 Replies Last Post 18 juin 2010, 03:50 UTC−4
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 16 juin 2010, 15:02 UTC−4
Hi

are you sure that you have an unique soliution in your case ?

Comsol converges only if you have a true BVP (boundary value problem) with an unique solution, else what to choose ?

COMSOL does not solve ALL ODE/PDEs just the class that has some physical sense, and we all still have our feet on ground (I hope so), we are not flaoting around via an arbitrary constant, no ?

Have fun Comsoling
Ivar
Hi are you sure that you have an unique soliution in your case ? Comsol converges only if you have a true BVP (boundary value problem) with an unique solution, else what to choose ? COMSOL does not solve ALL ODE/PDEs just the class that has some physical sense, and we all still have our feet on ground (I hope so), we are not flaoting around via an arbitrary constant, no ? Have fun Comsoling Ivar

Please login with a confirmed email address before reporting spam

Posted: 1 decade ago 17 juin 2010, 08:06 UTC−4
Hi Etienne,

as the electrical field is the gradient of the potential you can always add an arbitrary constant to your potential settings without changing the fields. However, as Ivar points out, you need to define the boundary conditions, otherwise the model gets meaningless.

I am doing a lot of magnetostatics and sometimes it gets a little dificult to set the boundary conditions at the outer boundary of the model to simulate the situation at infinty without making the model very big.

One workaround I found is the following: Sometimes my magnetic fields can be approximated by magnetic dipoles at larger distance. So I use the corresponding dipole field as boundary condition at the outer boundary. This allows to make the empty space to model much smaller.

So if you can approximate your charge distribution (maybe by a point charge) at larger distance, you could use the field (or potential) of the point charge at the outer boundary.

Cheers
Edgar
Hi Etienne, as the electrical field is the gradient of the potential you can always add an arbitrary constant to your potential settings without changing the fields. However, as Ivar points out, you need to define the boundary conditions, otherwise the model gets meaningless. I am doing a lot of magnetostatics and sometimes it gets a little dificult to set the boundary conditions at the outer boundary of the model to simulate the situation at infinty without making the model very big. One workaround I found is the following: Sometimes my magnetic fields can be approximated by magnetic dipoles at larger distance. So I use the corresponding dipole field as boundary condition at the outer boundary. This allows to make the empty space to model much smaller. So if you can approximate your charge distribution (maybe by a point charge) at larger distance, you could use the field (or potential) of the point charge at the outer boundary. Cheers Edgar

Please login with a confirmed email address before reporting spam

Posted: 1 decade ago 18 juin 2010, 03:50 UTC−4
Thank you Edgar and Ivar for your expertise.

I will try this charge "approximation".

Etienne
Thank you Edgar and Ivar for your expertise. I will try this charge "approximation". Etienne

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.