Re: [FEMM] Tape wound core modeling

[David Meeker <dmeeker@xxxxxxxxxxxxxxxxx>]

Dave Squires wrote:

Dave,
I think I asked you sometime back about the capability
to model tapewound cores. This would require, at minimum,
the ability to specify the direction of permeability at any angle.
Right now you can only use 0 or 90 degrees (x or y) in the
Lamination Attributes.  How hard would it be to at least add
in an arbitrary angle selection?  I have attached an example
of a rectangular tapewound core.  This is not very satisfying
as you have the 90 degree corners that don't really model
the curvature there.  If you had a round toroid
then it gets really bad as you can imagine.

It is easy to set PM magnetization direction by copying an arc
section around a circle.  It would be nice to be able to do
a similar thing with lamination direction to better model
tapewound cores.  So if you consider this seriously perhaps
work into it the ability to copy and rotate lamination direction
angles with the copy angle just like with PM magnetization direction.

Well, I haven't implemented a "on-edge" continuum lamination mode for AC problems--this sort of analysis only works for DC problems. I probably won't get around to making the DC on-edge lamination direction more general until such time as I address the AC case. The reason that I don't support this for AC problems is that I need to do more work to implement a continuum formulation that conserves eddy currents inside each lamination. The issue is eddy currents induced by flux entering in the direction normal to the lamination, an effect that doesn't appear in the "in-plane" lamination direction. (Also see http://femm.foster-miller.net/pipermail/femm/attachments/20021215/e6905b43/attachment.html) I'm interested in implementing a correct formulation for eddy currents in on-edge laminations (I'm a sucker for a good eddy current problem), but there are other improvements that are higher up on my list.  When I get to this one, it would make sense to define an aribrary lamination direction--this is probably something that I should have done in the original design of the program. 

Also, I would like to be able to plot BH curves of different materials
on top of each other for relative comparisons. At present it appears
there is no way to do that.  Could you add this ability to the BH
curve plotter?  I would prefer it to work by simply picking another
material from the library and allow loading of several or all non-linear
BH curves.

I didn't really intend the BH plotting tool to be used to compare various materials--the purpose is to make sure that the curve that FEMM fits through the BH data points looks sane enough.  To that end, I've implemented an optional log plot on the H axis in my development version of the program, such that it is easier to see what the curve is doing at low values of H.  At any rate, the material properties are stored in a human-readable text file such that BH curves could be plotted with some third-party plotting program like Gnuplot.

Dave.

-- 
David Meeker
Senior Engineer
Foster-Miller, Inc.
350 Second Avenue
Waltham, MA 02451-1196
781-684-4070
781-890-3489 (fax)
dmeeker@xxxxxxxxxxxxxxxxx