Note: This discussion is about an older version of the COMSOL Multiphysics® software. The information provided may be out of date.

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.

Eledtric Potential Boundary Condition in mef in V4.1

Please login with a confirmed email address before reporting spam

Hello Community,

I am using COMSOL Version 4.1.0.88, I got it yesterday. And there is someting very unfamiliar: in MEF I can only define the Boundary Condition Electrical Potential as a child feature to the BC Magnetic Insulation. And since that BC is only apllicable to the outermost model boundarys, I can only define Electric Potential at the outermost model boundarys. It seems to be impossible to define the electric potential of a boundary inside the model.

It was possible to do that in Versions 3.5 and 4.0a and it seems to be a really common need, so I wonder if I am missing something or if the COMSOL crew indeed changed that... And if they did - WHY?

I hope somebody can delight the situation.

Bye
MV

6 Replies Last Post 2010年12月9日 GMT-5 07:14
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 2010年12月7日 GMT-5 08:36
Hi

first of all get the latest update on the COMSOL web site as there is still a typo in one of the formulas fo "mef" in 4.1.0.88, and check again

see www.comsol.eu/support/updates/comsol41p/

--
Good luck
Ivar
Hi first of all get the latest update on the COMSOL web site as there is still a typo in one of the formulas fo "mef" in 4.1.0.88, and check again see http://www.comsol.eu/support/updates/comsol41p/ -- Good luck Ivar

Magnus Olsson COMSOL Employee

Please login with a confirmed email address before reporting spam

Posted: 1 decade ago 2010年12月7日 GMT-5 11:23
Hello,

We have indeed introduced limitations in how magnetic and electric boundary features can be combined. The reason is that certain combinations are unphysical and will result in ill-posed models that will be numerically unstable or give rise to unphysical solutions.

The Magnetic and Electric Fields interface is based on Maxwell-Ampère's law, in which the equation of continuity is inherent. Thus, any model must provide closed current loops. The current path can close as volume currents, surface currents, edge currents and combinations thereof.

Now, constraining the electric potential is normally associated with a current source or sink and is thus only allowed on a boundary supporting induced surface currents (magnetic insulation or impedance boundary condition). Magnetic insulation is by the way allowed on interior boundaries if you add it as a new feature. The default magnetic insulation feature is limited to external boundaries only.

Best regards,

Magnus Olsson
COMSOL


Hello, We have indeed introduced limitations in how magnetic and electric boundary features can be combined. The reason is that certain combinations are unphysical and will result in ill-posed models that will be numerically unstable or give rise to unphysical solutions. The Magnetic and Electric Fields interface is based on Maxwell-Ampère's law, in which the equation of continuity is inherent. Thus, any model must provide closed current loops. The current path can close as volume currents, surface currents, edge currents and combinations thereof. Now, constraining the electric potential is normally associated with a current source or sink and is thus only allowed on a boundary supporting induced surface currents (magnetic insulation or impedance boundary condition). Magnetic insulation is by the way allowed on interior boundaries if you add it as a new feature. The default magnetic insulation feature is limited to external boundaries only. Best regards, Magnus Olsson COMSOL

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 2010年12月7日 GMT-5 15:15
Hi

indeed it takes some time to convince oneself of the true effects of magnetic insulation/electric shielding BC's and the effective current loops.

That is also why following your ACDC course is very interesting.

Everything cannot be explained in the doc, I do partially agree, some features are of rather "natural physics knowledge" but are still not obvious before one have the occasion to play with physics as we can now with COMSOL

--
Good luck
Ivar
Hi indeed it takes some time to convince oneself of the true effects of magnetic insulation/electric shielding BC's and the effective current loops. That is also why following your ACDC course is very interesting. Everything cannot be explained in the doc, I do partially agree, some features are of rather "natural physics knowledge" but are still not obvious before one have the occasion to play with physics as we can now with COMSOL -- Good luck Ivar

Please login with a confirmed email address before reporting spam

Posted: 1 decade ago 2010年12月9日 GMT-5 02:27
Okay, thanks for the input, guys :)

With Magnus' hint I was able to model my problem the way I wanted to model it.
The old word popped into my mind: "It's not a bug - it's a feature!" ;)

Bye
MV
Okay, thanks for the input, guys :) With Magnus' hint I was able to model my problem the way I wanted to model it. The old word popped into my mind: "It's not a bug - it's a feature!" ;) Bye MV

Please login with a confirmed email address before reporting spam

Posted: 1 decade ago 2010年12月9日 GMT-5 04:19
I am using COMSOL 3.5a and want to solve a multi-phsics problem. One of the phenomenon I use is the the Meridional Induction and Electric Currents, Potentials. In the subdomain setting there are two vectoral diffrenetial equation. As far as I understand from What Magnus says, there is no need to define the current continuity equation. It is inherent in the vectoral differential equations. Am I right?

I am using COMSOL 3.5a and want to solve a multi-phsics problem. One of the phenomenon I use is the the Meridional Induction and Electric Currents, Potentials. In the subdomain setting there are two vectoral diffrenetial equation. As far as I understand from What Magnus says, there is no need to define the current continuity equation. It is inherent in the vectoral differential equations. Am I right?

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 2010年12月9日 GMT-5 07:14
Hi

I believe what Magnus refers to is that you need a closed path for your current in ACDC, typically if you i.e. simulate a 1/4 of a coil system you must ensure that the current can flow back via the appropriate BC's. This implies that not all BC's are compatible with this current flow

As well that in 3.5 this was not strictly imposed, which lead to illconditionned models with convergence issues, or not correct results. Getting a nice rainbow color view is easy, knowing for sure its correctly representing "reality" is another story (but the latter is the one you as engineer must stand up for)

--
Good luck
Ivar
Hi I believe what Magnus refers to is that you need a closed path for your current in ACDC, typically if you i.e. simulate a 1/4 of a coil system you must ensure that the current can flow back via the appropriate BC's. This implies that not all BC's are compatible with this current flow As well that in 3.5 this was not strictly imposed, which lead to illconditionned models with convergence issues, or not correct results. Getting a nice rainbow color view is easy, knowing for sure its correctly representing "reality" is another story (but the latter is the one you as engineer must stand up for) -- 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.