Hmm, okay no HTML except what yahoogroups adds...
>Craig, what do you think about individual laminated steel cores, or
>powdered iron or ferrite rod cores? Laminated steel would give me
>the ability to leave gaps for coolant passages built right into the
>core. I didn't realize such a large air gap was feasible with axial
>flux. Would say 1/4" offer any advantage over say 1/2", if the
>magnets didn't get hot? Are induced currents in a metal motor casing
>trivial, or would a nonmetallic casing of sufficient strength and
>rigidity be advantageous? I have been making a lot of assumptions
>all along, based on currently available motors and they all have
>steel or aluminum cases, but I figure I ought to ask.
The iron/alloy powder has very low losses because the particles
(which are insulated from each other) are too small to induce much
current into. The cores are toroids so you could theoretically run
coolant within, but in my admittedly limited experience so far, the
low losses seem to make anything beyond air cooling superfluous. The
low losses partly stem also from running low RPM s /frequencies.
Based on the manufacturer's data I calculated 1 W iron losses per
coil/core at 2000 RPM, or just 9W for a 4500(?) W motor. (I wish I
had more time for testing.)
One thing about the toroid cores and composites: they're not supposed
to get above 65 or 70 degrees celsius. You can't run it as hot as
with metal and motor varnish. The low heat generated is compensation.
Laminations are of course far better than solid iron, but they do get
a certain amount of stray currents induced into them. Ferrite -
oxides - isn't as good as actual metal, which the iron powder is as
well as the laminates. High RPM s might make ferrite worth using, but
my rotational frequencies (magnets passing by phases) are under about
100Hz.
If the gap is reduced, local attraction between magnets and iron
(powder) makes for rougher running, worse and worse as gap is
reduced. 3/8" is actually considered quite small. 1/2"+ seems to be
typical. Individual designs may vary of course and there may perhaps
be some that are quite different, but I know this is typical for
axial flux.
I found induced currents using a metal plate to mount the coils on
were definitely not trivial. Even tho the plate was over 1.5" from
the magnets it became very warm thanks to the extra electricity
needed to turn the rotor, and that would also apply more or less to
side walls. So the tough cast polypropylene-epoxy composite is
definitely advantageous. (PP is stronger and lighter than fiberglass,
pleasant to work with, and works well [torn into strips] in molds. I
think PP-epoxy is an overlooked composite - I don't use fiberglass
and I'll probably never bother with carbon fiber.)
The end wall behind the rotor (opposite side from the magnets & the
shaft end) could probably be metal with little adverse effect since
the flux is blocked by the rotor plate. Having switched to composite
bodies I haven't tried it yet.
Craig
http://www.TurquoiseEnergy.com
=====
Posted by: Craig Carmichael <craig@saers.com>
| Reply via web post | • | Reply to sender | • | Reply to group | • | Start a New Topic | • | Messages in this topic (6) |
No comments:
Post a Comment