Permeance coefficient calculation for magnets in FEMM 4.2

[flowwolf]

Dear Forumers,

This question is about load-line analysis for loudspeaker driver magnetic circuits in FEMM 4.2 (Finite Element Method for Magnetics), I can't seem to get an answer from the mailing lists.

I would like to get the permeance coefficient for random points inside the magnet.
I have problems interpreting the results reported by femm.
Most information sources tell me that pc = Bd/Hd, but B and H are vector quantities,
and I'm not sure which parts to use (either |B|,|H| or the components).

According to a JMAG newsletter I've read, Pc = - Bd / (µ0 Hd), where
"Bd [T], Hd [A/m]: The projection components in the magnetizing direction M of the magnetic flux density B and magnetic field H"

So does that mean that pc = Bz / Hz, if the magnetization direction is 90 degrees?I saw old mails about the permeance coefficient as well: http://www.femm.info/list/msg01827.html
where it was stated that Hd = H - Hc (if the "Demag. H in PMs" option is not set)

But subtracting Hc from |H| does not result in the |H| that is reported when "Demag. H in PMs" is setAll these confuses me, could I get some help about which values to use?

Also do Bd and Hd necessarily have to fall on the BH curve, or is the load-line given by
the intersection of a line with steepness of pc, where Bd and Hd are not necessarily on the BH curve?Any help would by appreciated.

Akos

 

[Paul Colby]

Are the magnetic materials isotropic (same in every direction)? If so the B and H vectors are in the same direction and I would think the magnitudes would be the natural choice. About the load line problem, I'd like to know more about it as I have similar modeling issues with piezoelectric materials. I too find it confusing.

 

[flowwolf]

Hello Paul,
I've received an answer from David Meeker:
"For your case, on a point-by-point basis, you'd want to compute:
Pc = (B.M)/(µ0*H.M)
where B is the reported B for a point inside a PM, H is the reported H for a point in the PM (with the default (checked) "Demag H in PMs" setting), and M is the magnetization in the block. Note that B, H, and M are all vectors, and . represents the dot product."

He pointed me to http://www.femm.info/Archives/misc/BarMagnet.pdf

 

[Paul Colby]

Very helpful, thanks. I once analyzed a variable reluctance motor using magnetic circuit theory. The parallel with piezoelectric materials should be understandable.