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 The information provided may be out of date.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.
Magntic field analyses using current excitation
Posted 13 mars 2013, 14:29 UTC−4 Low-Frequency Electromagnetics, Interfacing, Geometry, Modeling Tools & Definitions, Parameters, Variables, & Functions Version 4.3a 5 Replies
Please login with a confirmed email address before reporting spam
Hello,
I have started to use Comsol this week and have done some examples, but when I try to simulate my
model I hit some set backs.
The model has 16 electromagnetic poles with a pair of coils in quadrature wound on silicon steel.
So I have 32 coils each 90 turns.
I need to do the following:
1) Import the steel BH curve into Comsol for the Silicon Steel
2) Excite the coils using currents modeled by the following equations i.e.
ix = Ip*(cos(phi*pi/180)*cos(2*pi*freq*t)-r*sin(phi*pi/180)*sin(2*pi*freq*t));
iy = Ip*( sin(phi*pi/180)*cos(2*pi*freq*t)+r*cos(phi*pi/180)*sin(2*pi*freq*t));
Ip is the magnitude of the current phasor and r, freq are constants.
3) So should I choose stationary or time dependent for the study for the magnetic field physics?
4) Choose a method or physics for exciting the coils - is it Electrical Circuit(cir) or Magnetic Fields(mf)
5)The other question is how do I set the depth of my model
Generally my problem is similar to an assembly of inductors which produce flux similar to
a magnetic Halbach array.
Find the geometry attached (orange colored regions are coils, the grey regions are the
iron cored parts, the faint purplish color is defined as air).
I was able to import a 2D dxf file and I can also add the materials
(a) How can I view the domains and do I need to split the geometry to parts so that I can associate parts of the same material?
Regards and I hope you will assist me since am in the introductory stage.
Wanjiku
I have started to use Comsol this week and have done some examples, but when I try to simulate my
model I hit some set backs.
The model has 16 electromagnetic poles with a pair of coils in quadrature wound on silicon steel.
So I have 32 coils each 90 turns.
I need to do the following:
1) Import the steel BH curve into Comsol for the Silicon Steel
2) Excite the coils using currents modeled by the following equations i.e.
ix = Ip*(cos(phi*pi/180)*cos(2*pi*freq*t)-r*sin(phi*pi/180)*sin(2*pi*freq*t));
iy = Ip*( sin(phi*pi/180)*cos(2*pi*freq*t)+r*cos(phi*pi/180)*sin(2*pi*freq*t));
Ip is the magnitude of the current phasor and r, freq are constants.
3) So should I choose stationary or time dependent for the study for the magnetic field physics?
4) Choose a method or physics for exciting the coils - is it Electrical Circuit(cir) or Magnetic Fields(mf)
5)The other question is how do I set the depth of my model
Generally my problem is similar to an assembly of inductors which produce flux similar to
a magnetic Halbach array.
Find the geometry attached (orange colored regions are coils, the grey regions are the
iron cored parts, the faint purplish color is defined as air).
I was able to import a 2D dxf file and I can also add the materials
(a) How can I view the domains and do I need to split the geometry to parts so that I can associate parts of the same material?
Regards and I hope you will assist me since am in the introductory stage.
Wanjiku
Attachments:
5 Replies Last Post 14 mars 2013, 14:32 UTC−4
Please login with a confirmed email address before reporting spam
Posted:
1 decade ago
Hi
A challenging problem ;)
but if your signal is periodic, will it not be simpler to consider a frequency domain solver, you then define the amplitude and phase per coil sets ?
--
Good luck
Ivar
A challenging problem ;)
but if your signal is periodic, will it not be simpler to consider a frequency domain solver, you then define the amplitude and phase per coil sets ?
--
Good luck
Ivar
Please login with a confirmed email address before reporting spam
Posted:
1 decade ago
Thanks Ivar for your reply.
For now I have decided to use current density under Magnetic Fields physics, but there are some parameters I do not understand how they work:
Click on the Magnetic Fields(mf) on its window
a) I would like to know whether the thickness, d represents the depth of my model
b) By right-clicking the Magnetic Fields there is an option of Change Thickness(Out-of-Plane).
Can I use this to change the thickness of any domain in my model?
And if yes do I have to assign any boundary conditions at the interface with the rest of the models?
In addition is it centered, i.e. assuming my origin is (0,0,0), and my model is 20mm thick, lets assume I use the Change Thickness provision for a given domain and change it to 10mm, will the thickness be -5 to 5 with respect to the origin?
Regards
Wanjiku
For now I have decided to use current density under Magnetic Fields physics, but there are some parameters I do not understand how they work:
Click on the Magnetic Fields(mf) on its window
a) I would like to know whether the thickness, d represents the depth of my model
b) By right-clicking the Magnetic Fields there is an option of Change Thickness(Out-of-Plane).
Can I use this to change the thickness of any domain in my model?
And if yes do I have to assign any boundary conditions at the interface with the rest of the models?
In addition is it centered, i.e. assuming my origin is (0,0,0), and my model is 20mm thick, lets assume I use the Change Thickness provision for a given domain and change it to 10mm, will the thickness be -5 to 5 with respect to the origin?
Regards
Wanjiku
Please login with a confirmed email address before reporting spam
Posted:
1 decade ago
Hi
I understand you are in 2D, then the Z value is by default normally 1[m] (out of plane thickness = mf.d
But you can change some domains "Change thickness node"
Reminder, 2D means no gradients along Z hence no estimation of losses in Z direction neither, normally it's all PER meter depth.
A difference in depth does not take into account any coupling w.r.t. the difference of depth, it's only used for absolute calculations when integration on the "volume" is required (from my understanding, but there might eb exceptions, always check you understand well the COMSOL internal variables ;)
But 2D soles much quicker than 3D so it's worth the simplifications ;)
--
Good luck
Ivar
I understand you are in 2D, then the Z value is by default normally 1[m] (out of plane thickness = mf.d
But you can change some domains "Change thickness node"
Reminder, 2D means no gradients along Z hence no estimation of losses in Z direction neither, normally it's all PER meter depth.
A difference in depth does not take into account any coupling w.r.t. the difference of depth, it's only used for absolute calculations when integration on the "volume" is required (from my understanding, but there might eb exceptions, always check you understand well the COMSOL internal variables ;)
But 2D soles much quicker than 3D so it's worth the simplifications ;)
--
Good luck
Ivar
Please login with a confirmed email address before reporting spam
Posted:
1 decade ago
Hello,
I decided to try 2D first and if things work, I will proceed with 3D.
So there will be no effect in the distribution of magnetic field between the thick and the thin regions - I expect high flux-density in the thin regions. The solvers for MF are PDEs, and according to your statement below and assuming it is acting on 2D surfaces, i.e over areas not volumes, I assume it won't have any implication in my flux-density distribution irrespective of whether I change the thickness of a given domain using the Change Thickness node.
"A difference in depth does not take into account any coupling w.r.t. the difference of depth, it's only used for absolute calculations when integration on the "volume" is required"
I thought its a like quasi 3D analysis which would be a brilliant idea :)
Let me keep on trying I will get somewhere
Regards
Wanjiku
I decided to try 2D first and if things work, I will proceed with 3D.
So there will be no effect in the distribution of magnetic field between the thick and the thin regions - I expect high flux-density in the thin regions. The solvers for MF are PDEs, and according to your statement below and assuming it is acting on 2D surfaces, i.e over areas not volumes, I assume it won't have any implication in my flux-density distribution irrespective of whether I change the thickness of a given domain using the Change Thickness node.
"A difference in depth does not take into account any coupling w.r.t. the difference of depth, it's only used for absolute calculations when integration on the "volume" is required"
I thought its a like quasi 3D analysis which would be a brilliant idea :)
Let me keep on trying I will get somewhere
Regards
Wanjiku
Please login with a confirmed email address before reporting spam
Posted:
1 decade ago
Hi
Check the equation and see where "d" is used, and do a few test cases, then you will rapidly get a good feeling where the depth is important and where it can be ignored
--
Good luck
Ivar
Check the equation and see where "d" is used, and do a few test cases, then you will rapidly get a good feeling where the depth is important and where it can be ignored
--
Good luck
Ivar