Hastalloy.
Originally they were using silver and copper tapes. But as the fields exceed 20 tesla, the forces exceed the yield strength of the tape even in compression. As I recall, forces go way over 60kPsi. But that was about 5-10 years ago. And for obvious reasons, they abandoned silver as a substrate...($)
The real challenge is production of the tapes in industrial quantities. One issue is consistency of ultimate current capability along the kilometer lengths needed for production of magnets. It kills a magnet when a small area of the tape cannot support the design goal current in the magnetic field at temperature, and the only way to determine the quality of the tape is to measure it cold. And, to top it off, it is sensitive to bend radius because the superconducting film is on one side of the hastalloy. Since it is not located at the neutral plane of the conductor, bends with film on outside strains the film in tension, and strains in compression if bent the other way. I do not recall tape limits, but niobium 3 tin has big issues with .7% strain, totally wipes it's capability.
The small fusion groups (6 or 7 to date) are actually consuming all the HTS tape being produced.
Bill, the tape is very very expensive, even more than the niobium 3 tin conductors, a 3 foot diameter spool of that exceeds 250K dollars. I used to think my 2 dollar a foot 1mm niobium titanium cable was expensive, with one layer of a magnet using 1600 feet, 4 or 6 layers overall.
John
ps... mcmaster-carr...just ordered some 2 and 2.5mm screws and 1 meter of 2 and 3mm brass rod, order came to 30 something dollars, but then the invoice showed 27 dollars shipping...really?? Well, they've always been that way, my cost of doing business with them for fast turnaround.
Originally they were using silver and copper tapes. But as the fields exceed 20 tesla, the forces exceed the yield strength of the tape even in compression. As I recall, forces go way over 60kPsi. But that was about 5-10 years ago. And for obvious reasons, they abandoned silver as a substrate...($)
The real challenge is production of the tapes in industrial quantities. One issue is consistency of ultimate current capability along the kilometer lengths needed for production of magnets. It kills a magnet when a small area of the tape cannot support the design goal current in the magnetic field at temperature, and the only way to determine the quality of the tape is to measure it cold. And, to top it off, it is sensitive to bend radius because the superconducting film is on one side of the hastalloy. Since it is not located at the neutral plane of the conductor, bends with film on outside strains the film in tension, and strains in compression if bent the other way. I do not recall tape limits, but niobium 3 tin has big issues with .7% strain, totally wipes it's capability.
The small fusion groups (6 or 7 to date) are actually consuming all the HTS tape being produced.
Bill, the tape is very very expensive, even more than the niobium 3 tin conductors, a 3 foot diameter spool of that exceeds 250K dollars. I used to think my 2 dollar a foot 1mm niobium titanium cable was expensive, with one layer of a magnet using 1600 feet, 4 or 6 layers overall.
John
ps... mcmaster-carr...just ordered some 2 and 2.5mm screws and 1 meter of 2 and 3mm brass rod, order came to 30 something dollars, but then the invoice showed 27 dollars shipping...really?? Well, they've always been that way, my cost of doing business with them for fast turnaround.
Last edited:
It's the lorentz force. All magnet coils try to open up or expand when powered, we call it hoop forces when it's a solenoid. Think of the BLI product in a speaker, but make B 20 tesla, L as meters, and I as 1500 amps per square mm..
Remember that experiment where an insulated flexible wire loop is floating on a pool of mercury, and when current is put into the wire, it opens up to a circle.
Of course, back in the day before it became dangerous.
john
Remember that experiment where an insulated flexible wire loop is floating on a pool of mercury, and when current is put into the wire, it opens up to a circle.
Of course, back in the day before it became dangerous.
john
Last edited:
John,
Part of the cost is their deeeeep inventory...also, as you said their service.
When I had a machine down worth $500/hr if I ordered a part from McMaster Carr in the am in Burlington, NC, I could sometimes receive it via by 5 pm that same day out of their Atlanta warehouse...amazing service!
Watch out for those magnets, dude...
Howie
20T is not that bad, MRI goes up to 7T and you are directly exposed to all of it. AFAIK, the strongest magnetic field ever generated is some 45T (continuous) with a roadmap to 60T Magnet Sets World Record at 45.5 Teslas
Last edited:
I now find that if something breaks, I tend to buy more than one replacement. I hate down time, especially with my own toys. Explains why I have three 3-D printers...
Also explains why I've been building and printing parts bins, as I buy too many parts as spares...
The magnets that really scare me are the neodymium ones. They look just so...innocent. But if you get between them and a 3/4 inch wrench, they lose their innocence, you lose a finger.
At work, we have spares of darn near everything because lead times can be 6 weeks. Not one spare, as if you use it, you suddenly have none...so at least 3 to 5 for power supplies.
john
But something the size of a coke can, whilst super impressive science is not much use in a tokamak.
You can get 1200T provided you don't mind destroying the equipment and room it's in in the process! Full Page Reload
Limits with iron is roughly 1.2 to 1.4 tesla. A factor of 16 better at 20 tesla.
Problem is, saturation of the magnetic circuit, at that strength most of the field will not be directed by the iron, free to stray in space, not so much the gap.
The iron structures we use tend to lose permeability as the field goes up. As a magnet ramps up, the inductance starts high, but as the current rises, the inductance starts to fall, this due to the iron losing it.
jn
Ah, forgot to answer that.I dare not ask how you terminate the ends as the answer will probably melt my fragile brain, unless of course the answer is 'very very carefully'
Of course, very carefully...
Seriously, at these current levels, it's a simple overlap solder joint that is roughly 3 to 6 inches long, on the hastalloy side.
That other article 08 linked to talking about the insulation was weird..pretty much mixed up, some odd inaccuracies.
Tape wound solenoids can't use insulation because the insulation is plastic. It has too much give, and the hoop forces cause too much strain on the conductor. The expectation that the current can simply go around a "lattice defect" means doubling the current density on the turn adjoining it??? Not happening. That would mean running the coil at a design that is half the capability of the tape.
We call that a failure, and scrap it.
The second aspect of no insulation has two interesting qualities. First, it cannot be ramped up very fast, as turn to turn conduction will cause heating. However, during a quench, the rapid drop in current will cause the energy stored in the magnet to be distributed through the bulk of the superconducting pack, so there is less chance of a local overheating that can destroy the coil.
Protection of the coil through a quench is a very significant problem, as the energy stored in these puppies is in many cases the equivalent energy of the mass achieving orbit. We have to work hard to keep the maximum temperature below roughly 200C from absolute zero (approx), otherwise the solders will tend to melt. Luckily, HTS tends to quench propagate rather slowly. This because of the temperature, when you get to 50K all metals start to get their heat capacity back, so that tends to slow the heat propagation.
Niobium 3 Tin on the other hand, is extremely susceptible to quench damage as the current densities are high and the quench velocities low. We need to detect quenches of the conductor in the 10 to 50 millisecond regime, otherwise the coil will be toast. We have to extract as much energy from the magnet as possible, as these magnets are unable to survive the dissipation of all the magnetic energy stored. Sometimes we use distributed heaters to quench the entire coil to distribute the energy. These magnets, at 2 to 3 meg dollars per copy, are very very delicate and difficult to keep "undamaged" shall we say. If it were easy, we'd just buy them at Home Depot.
jn
Last edited:
Back in the day I was Engineering Director of Oxford Instruments. I was going to chime in about superconducting magnets, because we used to manufacture these every single day of the year in Nb:Ti and Nb:Sn. Both of these use a copper matrix with a large number of superconductor strands embedded. Even with state of the art construction you are limited to about 20 Tesla.
You can get higher than that, to around 100 Tesla in a pulsed magnet using very high purity rectangular section copper wire. You wind a solenoid, and shrink it into a very thick stainless steel tube. Cool to liquid nitrogen temperature at 77K and then shove current pulses at 100kA into it. In a concrete bunker just in case it blows up through hoop stress. We built quite a few of those, the first going to the National High Field Magnet Lab in Tallahassee.
Not as high as nature by a long stretch. One class of neutron star - called a magnetar - generates ~10^11 Tesla. A billion times greater than we can generate on Earth.
Nb:Sn is a bit of a black art. It is non-ductile, so the structure is that the Nb and the Sn are kept separate, so you can wind the magnet. The coil is then sintered to combine the two elements. And then it is either wax or epoxy impregnated.
But you cannot see a magnetic field. Even an MRI magnet, typically only 1 or 1.5 Tesla is dangerous. If you do the thought exercise of what happens if you let a piece of steel of around 1kg go from infinity in vacuum towards the magnet. It goes through the bore at about the speed of sound.
And yes indeed - the energy stored in a typical vanilla Nb:Ti magnet is enough to launch 1kg into orbit. Very scary bits of kit. Even at a trivial level, I've erased many credit cards by just walking through the test lab on my way to the parking lot.
You can get higher than that, to around 100 Tesla in a pulsed magnet using very high purity rectangular section copper wire. You wind a solenoid, and shrink it into a very thick stainless steel tube. Cool to liquid nitrogen temperature at 77K and then shove current pulses at 100kA into it. In a concrete bunker just in case it blows up through hoop stress. We built quite a few of those, the first going to the National High Field Magnet Lab in Tallahassee.
Not as high as nature by a long stretch. One class of neutron star - called a magnetar - generates ~10^11 Tesla. A billion times greater than we can generate on Earth.
Nb:Sn is a bit of a black art. It is non-ductile, so the structure is that the Nb and the Sn are kept separate, so you can wind the magnet. The coil is then sintered to combine the two elements. And then it is either wax or epoxy impregnated.
But you cannot see a magnetic field. Even an MRI magnet, typically only 1 or 1.5 Tesla is dangerous. If you do the thought exercise of what happens if you let a piece of steel of around 1kg go from infinity in vacuum towards the magnet. It goes through the bore at about the speed of sound.
And yes indeed - the energy stored in a typical vanilla Nb:Ti magnet is enough to launch 1kg into orbit. Very scary bits of kit. Even at a trivial level, I've erased many credit cards by just walking through the test lab on my way to the parking lot.
Last edited:
Never been brave/stupid enough to build one of those. One day....
Of course, the Japanese setup is the topfuel drag racing of the magnet world. A few seconds of glory then a pile of scrap.
High power magnets seems to be a continual fight with things trying to rapidly disassemble themselves. In the case of particle rings and tokomaks you are also trying to contain things that really don't want to be contained. Hats off to all the clever people who make sure that billion dollar oops moments are so rare!
Sawyers, you may have to explain "kits".. you also get caught up in the acronyms we use that normal humans have never seen.. ampere seconds is not a normal useage for most.
We use MIITS when we work the accelerator magnets, but that is a definition for energy deposited during a quench, not total energy of a magnet.
The scariest magnet for me was the 6 tesla solenoid with an 8 inch bore 2 meters long. So much energy, the fringe field from H##L. I'm not even involved in the testing of the Nb3Sn quads for LHC upgrade.. they also are scary at 16-18 kA.
It is weird when an inductor is 70 or 80 henries...and we are pushing hundreds of amps.
But from your resume, you've seen worse...
John
We use MIITS when we work the accelerator magnets, but that is a definition for energy deposited during a quench, not total energy of a magnet.
The scariest magnet for me was the 6 tesla solenoid with an 8 inch bore 2 meters long. So much energy, the fringe field from H##L. I'm not even involved in the testing of the Nb3Sn quads for LHC upgrade.. they also are scary at 16-18 kA.
It is weird when an inductor is 70 or 80 henries...and we are pushing hundreds of amps.
But from your resume, you've seen worse...
John
Last edited:
It is the thick end of 25 years since I worked for Oxford Instruments, and they have moved on since then. They now have specific fridges for Quantum Computing applications.
I was with the Research Instruments division. So our customers were research labs throughout the world. And standards labs, accelerator facilities etc. A more difficult customer base is difficult to think of!
And it wasn't just magnets. We made dilution fridges too, with a typical base temperature of 5mK (yup 5 thousandth of a Kelvin above absolute zero), which is far below the temperature in the known universe. And that was a standard product.
I was with the Research Instruments division. So our customers were research labs throughout the world. And standards labs, accelerator facilities etc. A more difficult customer base is difficult to think of!
And it wasn't just magnets. We made dilution fridges too, with a typical base temperature of 5mK (yup 5 thousandth of a Kelvin above absolute zero), which is far below the temperature in the known universe. And that was a standard product.