Re: [femm] Re: Approximation of non-axisymmetric geometry

[David Meeker <dmeeker@xxxxxxxx>]

jflrlk wrote:
>   Hi,
>
> I have a very similar situation, I would like to know the solution 
> too. 
>
> The motor we have is bascically axisymmetric, except that the 
> polarities of the surface mounted PMs around the rotor are 
> alternating, and there exists epoxy between every two adjacent PM 
> blocks. Does anybody know how I can model these PMs by an equivalent 
> axisymmetric case?
>
> Thanks a lot!
>From this description, it sounds one you can't really approximate this
one as axisymmetric. It sounds like more of a 2D planar problem, where
most of the flux flows in the plane of the alternating magnets, rather
than making axial loops--however, it's hard to say exactly how to
attack your problem without seeing what you are trying to model.

In general,  there could be an axisymmetric-like formulation (in the
sense of only having to mesh over r and z) in which the fields vary
sinusoidally in the tangential direction. This is a similar idea to
time-harmonic problems, in which the field is assumed to vary
sinusoidally with respect to time--the sinusoidal variation is just
with respect to a spatial direction, rather than with respect to time. 
However, FEMM doesn't solve that particular class of problem right now.

Dave.
-- 
David Meeker
dmeeker@xxxxxxxx
http://femm.berlios.de/dmeeker