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 The information provided may be out of date.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.
back induced voltage in a coil
Posted Feb 10, 2011, 3:44 p.m. EST Low-Frequency Electromagnetics Version 4.1 9 Replies
Please login with a confirmed email address before reporting spam
I am trying to calculate back induced coil voltage in a multiturn coil in both 2D axisymmetric and 3D geometries (same coil drawn in both geometries, models are included but need to be remeshed and resolved). However, the 2D axisymmetric model gives me the accurate value of real and imaginary component of Vind whereas the 3D model gives a different real component of Vind (Vind is calculated by integrating the coil electric field along the length of the coil).
9 Replies Last Post Feb 27, 2012, 12:20 p.m. EST
Please login with a confirmed email address before reporting spam
Posted:
1 decade ago
Hi
I have some issues with your models, when I load the 2D into my 4.1 I see a 2m radius for um coil Wires ? aren't there a unit issue in the geometry ?
in the 3D you have indeed a toroidal coil but where is the "air" around so the field may loop ?
--
Good luck
Ivar
I have some issues with your models, when I load the 2D into my 4.1 I see a 2m radius for um coil Wires ? aren't there a unit issue in the geometry ?
in the 3D you have indeed a toroidal coil but where is the "air" around so the field may loop ?
--
Good luck
Ivar
Please login with a confirmed email address before reporting spam
Posted:
1 decade ago
Hi Ivar,
There is an "air" block included around the 3D coil, I have just hidden this geometry domain for better post-processing visuals. As for the 2D model, that is the air block(semi-circle actually) that you are looking at which has a radius of 2m, zoom in to see the actual coil which is 0.01 m in radius.
There is an "air" block included around the 3D coil, I have just hidden this geometry domain for better post-processing visuals. As for the 2D model, that is the air block(semi-circle actually) that you are looking at which has a radius of 2m, zoom in to see the actual coil which is 0.01 m in radius.
Please login with a confirmed email address before reporting spam
Posted:
1 decade ago
Hi
yes I noticed for the 2D, but your wire diameter * number of turns was far smaller than the coil section (or did I miss something). I didnt get it to fit, anyhow 2m "air" is large, with an infinite element you can go smaller, with still good results
For the 3D I must have missed the air then ;)
--
Good luck
Ivar
yes I noticed for the 2D, but your wire diameter * number of turns was far smaller than the coil section (or did I miss something). I didnt get it to fit, anyhow 2m "air" is large, with an infinite element you can go smaller, with still good results
For the 3D I must have missed the air then ;)
--
Good luck
Ivar
Please login with a confirmed email address before reporting spam
Posted:
1 decade ago
Hi Ivar,
Yes, I agree that if you divide the cross section areas, you might get more than twice the number of turns. However if you include practical aspects such as a triple/quad insulation layer thickness, fill factor of 90%, then it matches with the number of turns. I did go overboard on the area for the 2D....I already had a previous geometry and was lazy enough not to mess with it, added the infinite element boundary on it too. I bet the results would not be any different, if I were to cut the area in half. Good observations as always Ivar.
Venkat
Yes, I agree that if you divide the cross section areas, you might get more than twice the number of turns. However if you include practical aspects such as a triple/quad insulation layer thickness, fill factor of 90%, then it matches with the number of turns. I did go overboard on the area for the 2D....I already had a previous geometry and was lazy enough not to mess with it, added the infinite element boundary on it too. I bet the results would not be any different, if I were to cut the area in half. Good observations as always Ivar.
Venkat
Please login with a confirmed email address before reporting spam
Posted:
1 decade ago
Hi
OK I have some more time to look at your models, you know you can clean and lighten the file with a File - Clear solution / clear mesh / reset model + Save. Because if you do a "clear mesh" at the mesh node it DELETES your MESH fully, inlcuding the pocedure to build it !
I see that you have a fill factor in 2D of about 40% indeed quite OK
In 3D I see the air now ;) though you have air and not Cu for your coil, I would use a coil conductance of about 6E7 S/m and not 1, that makes quite some difference. By the way you can define a cylindrical coordinate and use that to make the loop current equations far simpler
COMSOL does not warn you if there is no mesh, but you get an empty graphical screen without any data, quite normal, but unusual and not that obvious to see that your models are fully defined, except fr no mesh ;)
--
Good luck
Ivar
OK I have some more time to look at your models, you know you can clean and lighten the file with a File - Clear solution / clear mesh / reset model + Save. Because if you do a "clear mesh" at the mesh node it DELETES your MESH fully, inlcuding the pocedure to build it !
I see that you have a fill factor in 2D of about 40% indeed quite OK
In 3D I see the air now ;) though you have air and not Cu for your coil, I would use a coil conductance of about 6E7 S/m and not 1, that makes quite some difference. By the way you can define a cylindrical coordinate and use that to make the loop current equations far simpler
COMSOL does not warn you if there is no mesh, but you get an empty graphical screen without any data, quite normal, but unusual and not that obvious to see that your models are fully defined, except fr no mesh ;)
--
Good luck
Ivar
Please login with a confirmed email address before reporting spam
Posted:
1 decade ago
Hi Ivar,
I believe the feature of a model reset was in 3.5a but was for some reason discontinued in 4.0 when it first came out. I guess they have it back on again. I agree that it is good practise to be doing that.
I have a constant current excitation of the coil at 175 kHz and so do not have a copper coil conductance value.
I also work with a lot of non-cylindrical coils which are rectangular actually and so either have to use this way to define current directions or define them individually in seperate domains.
Venkat
I believe the feature of a model reset was in 3.5a but was for some reason discontinued in 4.0 when it first came out. I guess they have it back on again. I agree that it is good practise to be doing that.
I have a constant current excitation of the coil at 175 kHz and so do not have a copper coil conductance value.
I also work with a lot of non-cylindrical coils which are rectangular actually and so either have to use this way to define current directions or define them individually in seperate domains.
Venkat
Please login with a confirmed email address before reporting spam
Posted:
1 decade ago
Hi
I do not catch the reason why you have 1 S/m conductance, even if you apply a constant current. If you solve for V then V depends on the conductance and the current and there is a huge difference between 1 and 6E7 for the conductance, no ?
--
Good luck
Ivar
I do not catch the reason why you have 1 S/m conductance, even if you apply a constant current. If you solve for V then V depends on the conductance and the current and there is a huge difference between 1 and 6E7 for the conductance, no ?
--
Good luck
Ivar
Please login with a confirmed email address before reporting spam
Posted:
1 decade ago
Hi Ivar,
The way Vind is calculated is by integrating the electric field along the length of the coil. This method for the 3D model does give me accurate values for the imaginary part of Vind, which inturn gives me the accurate value of coil inductance. However, if I did indeed have coil conductance to be that of copper, then Vind, calculated in the above mentioned manner, real and imaginary, does not give me a value that is accurate.
Venkat
The way Vind is calculated is by integrating the electric field along the length of the coil. This method for the 3D model does give me accurate values for the imaginary part of Vind, which inturn gives me the accurate value of coil inductance. However, if I did indeed have coil conductance to be that of copper, then Vind, calculated in the above mentioned manner, real and imaginary, does not give me a value that is accurate.
Venkat
Please login with a confirmed email address before reporting spam
Posted:
1 decade ago
Hi Ivar,
The way Vind is calculated is by integrating the electric field along the length of the coil. This method for the 3D model does give me accurate values for the imaginary part of Vind, which inturn gives me the accurate value of coil inductance. However, if I did indeed have coil conductance to be that of copper, then Vind, calculated in the above mentioned manner, real and imaginary, does not give me a value that is accurate.
Venkat
Hi Venkat, I have a very similar problems when trying to model coils structure, just wondering you got the problem sorted out?
Regards
Chek