Re: [femm] femm 3.0 beta version available

[dcm3c@xxxxxxx]

In a message dated 5/17/00 8:13:46 AM Eastern Daylight Time, 
eckart.nipp@xxxxxxxx writes:

> Hi Dave,
> 
> finally I found a little time to try the new version of FEMM. Basically
> I liked the changes, especially the added tools for creating the geometry.
> Particularly useful is the possibility to define a group and multiple
> copy and rotate or translate it (e.g. slots in an electrical machine).
> Also the mirror tool is good but a nice feature would be the possibility
> to chose weather to keep or remove the original. Then mirroring in place
> can be carried out.

This would be easy enough to implement....

> So far I haven't discovered any bug but time was rather limited.
> 
> However, I wonder if I understood the circuit thing right. Or maybe it's
> just that I had a false expectation. What I thing is very useful is the
> possibility of coupling two or more domains and defining them as one
> circuit, for instance a short circuit winding where the induced current
> flows forward in one copper bar and back in another, i.e. on another
> location. Typical examples may be shaded pole motors and the cage rotor
> of asynchronous motors. Is it possible to obtain the induced currents in
> such parts using the circuit feature?

To do this sort of thing (like a shorted turn), you define a "circuit" in 
which the total current is zero. You then apply this "circuit" property to 
all the parts of your shorted turn (or cage rotor or whatever). This forces 
all of the currents in the part to add up to zero. To get the current in any 
one part, you can always do a "Total Current" block integral over the region 
of interest (e.g. only one part of a shorted turn, or in a particular bar of 
a cage rotor).

> In general I really like FEMM. It cannot (yet) compete with 10000 US$
> software packages but it's far better than what I thought was possible
> as freeware. Good work! (Only, I would prefer a linux based program, but
> that's another thing) :-)

On linux, you can always try MagSolve (http://elecmag3.ucd.ie/). I've also 
given out copies of the femm source in the past (and I am willing to in the 
future) for the purpose of ports to linux, but not much has come of it yet...

Dave.
--
http://members.aol.com/dcm3c