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.
Need precision on AC/DC module multi-turns coil
Posted 19 juil. 2017, 10:34 UTC−4 4 Replies
Please login with a confirmed email address before reporting spam
I'm using for the frist time the AC/DC module and particulary the multi-turns coil (the application is the design of a LVDT).
I have troubles understanding how comsol interpret the multi-turns coil.
On a first side I define a geometry, let say a 1cm by 10 cm rectangle (in 2D axisymmetric) for my coil and on the other side I define a multi-turn coil with the rectangle as domain, a number of spires and a wire gauge.
For me there is to much information over there and I would like to know how comsol handle this: If a define the number of turns and the gauge I have already a "geometry" length and width. What does comsol do with the extra information that is the selected domain, does it place the coils in the middle of the rectangle for example ? what if I define a rectangle not large enougth for the selected gauge ? How is defined the space between the spires ? How can I defined a multi layers coil?
thanks
Please login with a confirmed email address before reporting spam
let's put in some formal terms:
Let
A : area of the rectangle
N: number of turns
g : gauge ie cross section of the conductor
First the area A is not only filled with conductor but also there is some
insulation. We can define a conductor fill factor as
cff = N*g/A
For something reasonable this value must be smaller that 1 (say 0.6).
To my knowledge it is not check be Comsol and it is up to the user to enter
reasonable values. In my opinion the behaviour of Comsol is ok (and you
may have a different point of view).
Indeed these values do not really enter the finite element computation and
are only for convenience of input and output.
If you prescribe a current I, Comsol will apply an external current density
N*I/A on the entire rectangle. Only for computing the Ohmic resistance
(which is an analytic formula and no finite elements involved) the wire cross
section is needed.
I advise you to read the respective chapter in the manual. Multiturn coils
are a bit more complex that it may seem. For example they assume the all wires
are equally distributed over the area, and don't care about the number of layers.
Regards
Jens
Please login with a confirmed email address before reporting spam
Thank you very much for your reply !
I think I better understand how Comsol manage its multi turn coil.
Let's say that I want to modelize the "real" coil which the cross section is on the left side of this picture : gofile.me/2NLgz/J7BvFojds . (blue is insulating material and yellow conducting)
What is best :
-Use a rectangle geometry with a W width and a L length.
-Use a rectangle with the "equivalent conducting aera" of my coil (8*S1 here) keeping the same W/L ratio for the rectangle.
Since comsol will define the current density on the entire rectangle the second approch seems more appropriate for me.
Thanks again.
Please login with a confirmed email address before reporting spam
I would use the option:
-Use a rectangle geometry with a W width and a L length.
With just eight turns you might consider using the single turn feature.
Is this a transient or frequency study? Note that the multi-turn approach
is only valid if the skin depth is larger than the dimensions of the wire.
Regards
Jens
Please login with a confirmed email address before reporting spam
Hello, I am currently workin on a 3-D comsol 5.3a in AC/DC to generate the quasistatic magnetic field of order 65kA/m with frequency of 0.01Hz by the use of multiturn coil in which input is current but that thus not generate the required magnetic field. So can anyone suggesrt any method to generate the magnetic field.