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.

Rotation of the model

Please login with a confirmed email address before reporting spam

Hello
I am trying to simulate electromagnet laying on "X" axis.Geomtry defiined as cylinder.
The problem that i have, is to deffine external current density ,because the magnet is rotated 30 degrees about "Y" axis .
I deffined rotated coordinate system with beta = pi/6.
External current density deffined : x=0; y= J0*z/sqrt(y^2+z^2); z= J0*y/sqrt(y^2+z^2); with rotated coordinate system.
In the result i see that current still not rotated.

What i did wrong or am i missing some thing?

Thanks in advance

Dimitri

9 Replies Last Post 28 juil. 2010, 03:42 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 24 juil. 2010, 03:23 UTC−4
Hi

if you are in V4 it's rather easy to define a cylindrical coordinate system and define your current density (0,J0,0) for (r,phy,a). You have plot options to "see" the orientation of the coordinate system (see Results)

you can also define a variable ryz=sqrt(y^2+z^2) (for your results note that you might be better out with Y,Z of the material frame and not y,z of the spatial frame)

And for angles double check the units required, there are fields in [rads] and in [deg] (it's normally specified, but easy to get wrong).

Now for your formula, if applied to the right coordinate system, and that J0[A/m^2] starts at y=0,z=0 it's probably right but I'm not 100% sure as I still do not fully visualise your model ;)

--
Good luck
Ivar
Hi if you are in V4 it's rather easy to define a cylindrical coordinate system and define your current density (0,J0,0) for (r,phy,a). You have plot options to "see" the orientation of the coordinate system (see Results) you can also define a variable ryz=sqrt(y^2+z^2) (for your results note that you might be better out with Y,Z of the material frame and not y,z of the spatial frame) And for angles double check the units required, there are fields in [rads] and in [deg] (it's normally specified, but easy to get wrong). Now for your formula, if applied to the right coordinate system, and that J0[A/m^2] starts at y=0,z=0 it's probably right but I'm not 100% sure as I still do not fully visualise your model ;) -- Good luck Ivar

Please login with a confirmed email address before reporting spam

Posted: 1 decade ago 24 juil. 2010, 07:34 UTC−4
Hi
Thanks for the answer :)

but i still stuck....

I defined cylindrical coordinate system , and as you explained "Je=(0;J0;0).... but i got current only on Y axis and not circular shape.

Its seems like it still use global coordinate system.

How can i to see both coordinate systems on the plot?


Picture attached

Thanks again
Dimitri
Hi Thanks for the answer :) but i still stuck.... I defined cylindrical coordinate system , and as you explained "Je=(0;J0;0).... but i got current only on Y axis and not circular shape. Its seems like it still use global coordinate system. How can i to see both coordinate systems on the plot? Picture attached Thanks again Dimitri


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 24 juil. 2010, 07:44 UTC−4
Hi

Are you shure you have selected the right coordinate system when you entered (0,J0,0) because it looks indeed that you are using the default one. in V4 you must define it under Model>Definitions>Coordinate system, and then when applying BC's you must select the coordinate system to use from the selection list. In 3.5 its basically the same, but the definition of the CS is elsewere (I do not have access to COMSOL from my current holiday residence ;)

--
Good luck
Ivar
Hi Are you shure you have selected the right coordinate system when you entered (0,J0,0) because it looks indeed that you are using the default one. in V4 you must define it under Model>Definitions>Coordinate system, and then when applying BC's you must select the coordinate system to use from the selection list. In 3.5 its basically the same, but the definition of the CS is elsewere (I do not have access to COMSOL from my current holiday residence ;) -- Good luck Ivar

Please login with a confirmed email address before reporting spam

Posted: 1 decade ago 24 juil. 2010, 10:08 UTC−4
Thanks again....

i think i have some problem with the software :(

I simplified the problem, the magnet just laying on X axis and still its not working....

Please check the pictures , its the steps as i deffined ...


I am at holiday residence too bust still have to finish it ....
Thanks again.... i think i have some problem with the software :( I simplified the problem, the magnet just laying on X axis and still its not working.... Please check the pictures , its the steps as i deffined ... I am at holiday residence too bust still have to finish it ....


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 24 juil. 2010, 11:19 UTC−4
Hi again

Well from the two first images I have nothing to say, from the last one I see a suspicious item in your formula:

y component (also by similarity for the others):
mf.By + sys2.phi + mf.Jy

I would have expected a transform value for the sys2.phi to x,y and z and not the raw sys2.phi value.

Normally there is a way to get the inverse transform of a coordinate change. But as my holiday demo version is out on date, I cannot check how this is now.

Have you tried to use the keyboard shortcut CNTR+space to get the variable name lists ? Normally the built in variables take care of the coordinate change. There is also the Show Equation and checking of how COMSOL fills in the current density as it should normally get the cartesian values from the coordinate change

Have you tried to integrate the current over a coil section, you should get out J0*area, as check ?

Also there were some additional options in the plot (right-click) at least in V4.0.993 (latest update) linked to coordinate systems, didnt manange to test them

--
Good luck
Ivar
Hi again Well from the two first images I have nothing to say, from the last one I see a suspicious item in your formula: y component (also by similarity for the others): mf.By + sys2.phi + mf.Jy I would have expected a transform value for the sys2.phi to x,y and z and not the raw sys2.phi value. Normally there is a way to get the inverse transform of a coordinate change. But as my holiday demo version is out on date, I cannot check how this is now. Have you tried to use the keyboard shortcut CNTR+space to get the variable name lists ? Normally the built in variables take care of the coordinate change. There is also the Show Equation and checking of how COMSOL fills in the current density as it should normally get the cartesian values from the coordinate change Have you tried to integrate the current over a coil section, you should get out J0*area, as check ? Also there were some additional options in the plot (right-click) at least in V4.0.993 (latest update) linked to coordinate systems, didnt manange to test them -- Good luck Ivar

Please login with a confirmed email address before reporting spam

Posted: 1 decade ago 26 juil. 2010, 09:46 UTC−4
Hi Ivar
I tried all tips that you gave, still have the problem ....

I have v40 V4.0/0.937 and any shortcuts not work

How can i to solve it?

thanks again

Dimitri
Hi Ivar I tried all tips that you gave, still have the problem .... I have v40 V4.0/0.937 and any shortcuts not work How can i to solve it? thanks again Dimitri

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 26 juil. 2010, 11:21 UTC−4
Hi

shortcuts are just handy, the latest 4.0a update patch is on the main website, everything else should already be in your 0.937, it could be that if your formula got directly this .phi value then there is still a little bug in your version, probably updated and corrected since, try to get a v4.0a CD and update to 4.0.993. It's more stable

--
Good luck
Ivar
Hi shortcuts are just handy, the latest 4.0a update patch is on the main website, everything else should already be in your 0.937, it could be that if your formula got directly this .phi value then there is still a little bug in your version, probably updated and corrected since, try to get a v4.0a CD and update to 4.0.993. It's more stable -- Good luck Ivar

Please login with a confirmed email address before reporting spam

Posted: 1 decade ago 26 juil. 2010, 15:28 UTC−4
Hi

Thanks

I will try to get 40a

Meanwhile i reinstalled comsol, so now something else occured, when i am trying to defiine Cylindrical coordinates and i choose material frame, its cant be choosen to be coordinate system ( it not appear in the selection window )....but if i choose spatial frame the it seems ok, but still wrong directions :(

See picture attached

Can you please to give me example how i should deffine it?

Thanks in advance

Dimitri
Hi Thanks I will try to get 40a Meanwhile i reinstalled comsol, so now something else occured, when i am trying to defiine Cylindrical coordinates and i choose material frame, its cant be choosen to be coordinate system ( it not appear in the selection window )....but if i choose spatial frame the it seems ok, but still wrong directions :( See picture attached Can you please to give me example how i should deffine it? Thanks in advance Dimitri


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 28 juil. 2010, 03:42 UTC−4
Hi

turn on the equations: "options properties" and check if you can get more info about the underlaying variables, it helps often to better understand how Comsol does things

--
Good luck
Ivar
Hi turn on the equations: "options properties" and check if you can get more info about the underlaying variables, it helps often to better understand how Comsol does things -- 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.