Copper Electromagnets and Power and Heating limits

Discuss the technical details of an "open source" community-driven design of a polywell reactor.

Moderators: tonybarry, MSimon

RobL
Posts: 35
Joined: Thu Feb 24, 2011 4:14 pm

Post by RobL »

For cheaper testing purposes liquid helium cooled pulsed coils make a lot of sense, at 10K it appears that very pure copper can have less than 0.01% of room temp conductivity:
http://indico.cern.ch/getFile.py/access ... nfId=90787
And it is a lot cheaper and easier to fabricate.

Aluminium is not quite so good, with slightly higher resistivity, thought the gains are still huge.

Downside is of course pulsed power supplies needed to quickly ramp up magnets, but the magnets themselves would be cheap to make

happyjack27
Posts: 1439
Joined: Wed Jul 14, 2010 5:27 pm

Post by happyjack27 »

off the top of my head i see two ways to cool it.

- assuming superconductor -

w/sc you only have to do external heat. so there's really no point in putting a cooling tube through the center. that just tells you your outermost layer is gonna get fried before the center even gets warm. you put the coolant, on on the outside, on the other hand, and all of the coils are gonna stay the same temperature (as there's no internal heating), so you're cooling the coils perfectly evenly.

so there, there's method 1. a thin channel of liquid nitrogen or whatever circuilating on the outside of the coil. you pump it through the anti-nubs into one of those heat pump thingies.

method 2 is a bit more minimal, and it only works if method 1 is overkill. don't even bother to pump it around the coil. the wire or casing itself should conduct heat well enough to transfer it to the anti-nubs fast enough. then you just cool at the anti-nubs. and maybe you can use e.g. phase-change cooling to transfer heat away faster at the anti-nubs. assuming you don't get it _too_ cold and crack the wires.

one or both of those methods i would think would work fine with minimal impact on the physical processes or other engineering concerns.

RobL
Posts: 35
Joined: Thu Feb 24, 2011 4:14 pm

Post by RobL »

Could you get away with using toroidal billets of superconductor to eliminate fabrication problems? Put those billets inside insulated cryostats. While warm insert them axially between two high field helmholtz coil solenoids to create a large flux through toroid bore. Cool cyrostat below superconducting transition temp and then pull it out of the helmholtz field.

Should now have huge current induced in toroidal billet that (at least theoretically) won't dissipate as long as kept below transition temp.

Assemble 6 of these into polywell.

KitemanSA
Posts: 6179
Joined: Sun Sep 28, 2008 3:05 pm
Location: OlyPen WA

Post by KitemanSA »

Interesting. The topic is LN cooled Copper and everyone is talking superconductors.

Tom Ligon
Posts: 1871
Joined: Wed Aug 22, 2007 1:23 am
Location: Northern Virginia
Contact:

Post by Tom Ligon »

Kiteman,

Sorry to not get back to you. I had been searching for LN2-cooled electromagnets and the first hit was to this discussion. I was just doing a little preliminary fooling around.

At the moment my hope is to try some basic experiments on the practicality of building a magrid this way, way short of building a working Polywell. The little chamber I use for my demo fusor is way too small and it lacks high-vacuum pumping. A good enough chamber plus power supplies to do this would get into the tens of thousands of dollars, and is beyond my immediate ambitions (but maybe one day).

Closer to practical for me is working out how to build electromagnets. If nothing else, I can certainly dunk copper windings into LN2. Working out the support, insulation, and casings to make a practical coil ought to be within my means.

Now buried on my desk is an older study in which the resistivity of copper at cryo temperatures is investigated ... the numbers did not at first blush look as good at LN2 temperatures as I thought I had heard. It looked like the resistance drops as a fairly straight line function of temperature until you get down near absolute zero. On top of that, what real experiments need is data on real materials ... what can we do with available copper wire, strip, and tubing?

My latest thinking is that the magnets I could build might be better done with copper wire. It can pack to a round cross section (ribbon wants to go into square). Electrical wire tends to run about 99.99% pure copper, as compared to who-knows-what for tubing. Tubing cooled by LN2 flow would have the same instability problems as water-cooled, which is that any point at which the fluid boils becomes a pressure-generating hot spot pretty quickly. So flooding a casing packed with wire may be the most practical way of making an LN2-cooled electromagnet (adaptable maybe to LH2 once proved out) for practical lab work. I'm thinking in terms of long pulses, maybe seconds to a minute of run time.

There are some overbraids capable of LN2 temperature with which to insulate the wire. Fiberglass braid does not seem to be rated below about -70 C.

Tom Ligon
Posts: 1871
Joined: Wed Aug 22, 2007 1:23 am
Location: Northern Virginia
Contact:

Post by Tom Ligon »

This paper is almost as old as I am. But available online, and I don't think copper is much different now than then.

The Thermal and Electrical Conductivity of Copper at Low Temperatures

White, G. K.

Australian Journal of Physics, vol. 6, p.397

Bibliographic Code: 1953AuJPh...6..397W


http://adsabs.harvard.edu/full/1953AuJPh...6..397W

Tom Ligon
Posts: 1871
Joined: Wed Aug 22, 2007 1:23 am
Location: Northern Virginia
Contact:

Post by Tom Ligon »

Some of this material got me curious about LH2, which I have never worked with before. It is pretty interesting stuff once you get past the Hindenburg problem.

The boiling point at one atmosphere is way colder than LN2, and not all that much higher than liquid He. Liquid helium is cold, but has absolutely lousy heat of vaporization. LH2 looks like it will carry off twice the heat as LN2.

LN2 is fairly readily available around my neck of the woods (for instance, the huge tank of it out back that automatically calls in an order from a tanker truck when it gets low). LOX is generally available almost everywhere due to medical uses. I've never checked for the local availability of LH2. Maybe I should. A quick looks shows an interest in LH2 as a motor fuel, widespread adoption of which would make it easy to come by. At least one firm makes a box about the size of a chest freezer that makes and liquifies hydrogen for motor fuel use.

http://www.engineeringtoolbox.com/cryog ... d_492.html

http://www.hydrogencarsnow.com/liquid-hydrogen.htm

NewtonPulsifer
Posts: 3
Joined: Thu May 05, 2011 4:19 am
Location: California

Post by NewtonPulsifer »

There's a small company with a carbon nanotube impregnated aluminum alloy (also some brass etc in there - google the patent if you want - Bourque Industries).

It exceeds Copper 11000 in thermal and electrical conductivity. Nice. Normally aluminum is about 59% of copper in thermal and electrical conductivity.

"The most popular form of pure copper is the standard electrical wire grade of copper (C11000) contains 99.95% Cu, 0.03% O2, and less than 50 ppm metallic impurities."

Kryron is also stopping bullets nicely in tests so far.

Maybe sometime in the near future carbon nanotubes plus a copper alloy will be able to cheaply double copper's performance.

Sources:

Info on Copper 11000 http://www.copper.org/resources/propert ... ppers.html.

Link to article about GE in Tucson testing this new "Kryron" alloy vs. Copper http://money.msn.com/business-news/arti ... d=13767743

EDIT: rather than pollute this thread we can talk about Kryron here; I opened a topic in the News section : viewtopic.php?t=3275

Post Reply