GIThruster wrote:IIRC, the National High Magnetic Field lab down at FL State is working with Superpower Inc for some years now, using the latter's 2nd generation high temperature superconductor tape. The stuff has a nice tight bend radius (one of the most important breakthroughs so far as use in windings) and carries fantastical amounts of current. Single turn windings with 4,000+ amps and almost 20T fields generated. The size and shape of mobile power generators like what is used aboard ships and the size of electric motors are both going to shrink fantastically. And especially when you have a motive application, this is a big deal.
To expand on this and perhaps expose my ignorance., WB6 had ~ 100,000 amp turns for a 0.1 Tesla field, Mini B had ~ 40,000 amp turns for almost 0.3 T. My unbderstanding is that the B field at the cusp facing surface of a round electromagnet is proportional to the radius of the magnet. Thus Mini B with ~ 1/2 the radius of WB6 had 4 times the B field strength, assuming the same number of amp turns. With ~ 1/2 the number of amp turns the resultant strength was ~ twice that of WB 6, close enough for comparison, especially as the B field strength of WB6 was a moving target. Values of ~ 0.1 to 0.2 teslas was given as the capacity, depending on how much amps they pushed through it. If this superconductor with one winding gives a B field of ~ 20T with 4,000 amp turns, that suggests a radius of ~ WB6 B field *100,000 amp turns= Quoted B field *4000 amp turns / R^2 .
With ~ 1/ 20th the amp turns, the radius would have to be smaller by a factor of square root of the radius, or about 15 cm ^2 / x cm^2= 20.
This is ~ 225/ 11=20. sqr of 11 is ~ 3.3 or 3.3 cm radius of curvature for the same B field strength, but at 100 times greater B field I think the corresponding radius of curvature would be ~ 0.033 cm or 0.3 mm.
That is a pretty small radius of curvature. If true, the wire would obviously need to have a diameter much smaller than this. The current carrying capacity of supperconductors is generally given as the current carrying capacity with a cross section of one cm^2. A corresponding superconductor ribbon much smaller than 1 mm in cross section would carry ~ 10 Amps.
The numbers do not add up. Either my analysis is screwed up or something is missing. A 20 T field may be possible with many windings instead of just one, and this is not apparent in the quote. This would be consistent with ~ 20-30 T fields being generated in modest sized electromagnets with many turns. These are, I believe, cold superconductors, or even combinations of copper conductors and superconductors.
The capacity of ~ 4,000 amps/ cm^2 seems to be a common capacity for several high temperature superconductors. I have seen up to ~ 8000 Amps/ cm^2 for a cold/ liquid helium cooled super conductor.
If a superconductor can handle 4000 Amps / cm^2, and 100,000 amp turns is needed for a 15 cm radius magrid, then 25 turns or 25 cm^2 of cross section is needed. With upscaling the radius to 1-2 meters, the number of amp turns would need to go up the square of this proportion. Given that the machines are probably going to be at least several meters in diameter/ greater than 1 meter in radius, the engineering considerations are probably more concerned with the cross section of the superconductor ribbons necessary for the chosen current, and the cooling concerns to manage deep penetrating x- rays and possibly neutrons are the biggest challenges. The superconducting magnets in the LHC are possibly representative of what will be be needed for a polywell. The difference would be the additional shielding and cooling to handle the external/ penetrating thermal load. The LHC uses cold superconductors, and I suspect the tokamak plans call for this. If high temperature superconductors could be used, the cooling concerns would be improved some. The Polywell has internal magnets exposed to the plasma to a degree (through ExB diffusion) and also any x-rays or neutrons. The LHC does not have this challenge. And the tokamac has the magnets outside the first wall and cooling blanket of the reactor so it also is less challenging.
So long as the radius of curvature of the superconductor ribbon is less than a meter or so , it is probably not a concern. Considerations that can ease the situation is probably the improvement in current carrying capacity of the ribbon and the operating temperature. Liquid nitrogen temperatures are better, room temperature would be great.
What is the current carrying capacity of copper wire? And, how much cooling can you achieve with vigorous coolent flow? Ohmic heating is only a limit once you pass this cooling capacity. Also, of course, ohmic heating will consume megawatts, or even hundreds of Megawatts of power, and superconductors avoid this complication. The final question is what are the best compromises between the two approaches.
PS: Just to complicate the issue further. If you operate at liquid nitrogen temperatures the conductance of simple (high purity) copper wire is 7-8 time better than at room temperature. At 4 degrees K the conductivity is ~ 20 times better. As such, the advantages of superconductors versus the complications, become more narrow.
Dan Tibbets