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 The information provided may be out of date.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.

Rotational Symmetry in Model -> Memory Saving Model (Electromagnetics)

Please login with a confirmed email address before reporting spam

Hi all!

I am simulating the force on an iron piece caused by the magnetic field which is inducted by a coil. While simulating, my workstation is running out of memory (Win XP 32 bit, 3 GB memory). I am thinking about buying more memory and installing a 64 bit OS. But besides I would like to create a memory saving model.

The model consists of a rotational symmetry (see full-model.mph). I tried to simulate one quarter and one half of the model but I didn’t find the right boundaries (see half-model-mph).

Is it possible to reduce the model to one half, a quarter or even less? Which boundaries and conditions are necessary for the simulation?

It would be great if you could help me out!

Thanks!
Philipp


6 Replies Last Post 16 dic 2010, 04:35 GMT-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 13 dic 2010, 08:32 GMT-5
Hi

1) first of all use some Cu for the coil, with a conductivit of 0S/m I'm not sure the model is easy to solve
2) what about starting in 2D axi it's much quicker to solve, and you have more advanced BC's for multi turn coils, just do not forget that the "loop length" is 2*pi*r and check if the 2*pi si with or not (things have changed betwenn 3.5 and 4.0a but I still havent udnerstood if 4.1 is back at 3.5a convention without the 2*pi ;)

--
Good luck
Ivar
Hi 1) first of all use some Cu for the coil, with a conductivit of 0S/m I'm not sure the model is easy to solve 2) what about starting in 2D axi it's much quicker to solve, and you have more advanced BC's for multi turn coils, just do not forget that the "loop length" is 2*pi*r and check if the 2*pi si with or not (things have changed betwenn 3.5 and 4.0a but I still havent udnerstood if 4.1 is back at 3.5a convention without the 2*pi ;) -- Good luck Ivar

Please login with a confirmed email address before reporting spam

Posted: 1 decade ago 14 dic 2010, 03:57 GMT-5
Hi Ivar,

thanks for your reply. I would prefer using a 3D simulation because I use Solidworks for modeling and the LiveLink for Solidworks to make quick geometry changes.

I could use a 2D simulation if there is a possibility to break down 3D to 2D in Comsol. Maybe by using work planes?

Thanks again!
Philipp
Hi Ivar, thanks for your reply. I would prefer using a 3D simulation because I use Solidworks for modeling and the LiveLink for Solidworks to make quick geometry changes. I could use a 2D simulation if there is a possibility to break down 3D to 2D in Comsol. Maybe by using work planes? Thanks again! Philipp

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 14 dic 2010, 04:49 GMT-5
Hi

unfortunately COMSOL havent yet understood the interest to do 2D from SW CAD, so I manage so far only to export 2D (sketches or selected faces) by playing with the different formats (I cannot rememer now if it was via x_T or step, but one of these allowed me to export 2D shapes rapidly from SW (2010), anyhow DXF export works always

Neither has COMSOL allowed to define a "cut plane" for a 3D geometry to export a projection as a 2D geometry to be used in 2D or 2Daxi physics, they show you the projection, but you cannot import the edges as you can in SW, you must overwrite by hand. This is a pity, but if we are many to "complain" perhaps they will consider it with higher priority ;)

2D axi solves much quicker and has more advanced BC for coils, so it's worth the effort to find a quick way around

--
Good luck
Ivar
Hi unfortunately COMSOL havent yet understood the interest to do 2D from SW CAD, so I manage so far only to export 2D (sketches or selected faces) by playing with the different formats (I cannot rememer now if it was via x_T or step, but one of these allowed me to export 2D shapes rapidly from SW (2010), anyhow DXF export works always Neither has COMSOL allowed to define a "cut plane" for a 3D geometry to export a projection as a 2D geometry to be used in 2D or 2Daxi physics, they show you the projection, but you cannot import the edges as you can in SW, you must overwrite by hand. This is a pity, but if we are many to "complain" perhaps they will consider it with higher priority ;) 2D axi solves much quicker and has more advanced BC for coils, so it's worth the effort to find a quick way around -- Good luck Ivar

Please login with a confirmed email address before reporting spam

Posted: 1 decade ago 15 dic 2010, 03:57 GMT-5
Hi,

that's a pity. I could really use a way to convert 3D to 2D using Comsol ...

Anyway, I will check some 2D simulations hoping that the computation time and the memory use improves.

Thanks for your help, Ivar!
Philipp
Hi, that's a pity. I could really use a way to convert 3D to 2D using Comsol ... Anyway, I will check some 2D simulations hoping that the computation time and the memory use improves. Thanks for your help, Ivar! Philipp

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 15 dic 2010, 04:50 GMT-5
Hi

2D givs a great reduction in solving time ;)

Tody the only way around is to cut your 3D by a workplane an then redraw in 2D over the projecton image. This is OK for a simple model, but you can forget it for most typical engineering models with houndred of domains. Only a correct *D cut plane geoemtrical function can do that, in the parasolid kernel this should be there, but I'm not sure its in the list of what is already licenced to COMSOL

--
Good luck
Ivar
Hi 2D givs a great reduction in solving time ;) Tody the only way around is to cut your 3D by a workplane an then redraw in 2D over the projecton image. This is OK for a simple model, but you can forget it for most typical engineering models with houndred of domains. Only a correct *D cut plane geoemtrical function can do that, in the parasolid kernel this should be there, but I'm not sure its in the list of what is already licenced to COMSOL -- Good luck Ivar

Please login with a confirmed email address before reporting spam

Posted: 1 decade ago 16 dic 2010, 04:35 GMT-5
Hi again,

I made a 2D axi simulation of the problem. But there must be something wrong and I cannot find what it is:

The electromagnetic force on the slider in r-direction should be zero (or at least almost zero). But it isn't! See "force-2D.mph".

Did I forget something?

Philipp
Hi again, I made a 2D axi simulation of the problem. But there must be something wrong and I cannot find what it is: The electromagnetic force on the slider in r-direction should be zero (or at least almost zero). But it isn't! See "force-2D.mph". Did I forget something? Philipp

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.