Hacker News Reader

Comment by fbdab103 | Hacker News Reader

Let's say there was a superconducting material that operated at atmospheric pressure and 25C. While waving the magic wand, let us also pretend it costs $1000/kg.

What are the immediate applications for such a thing? I can trivially imagine long distance electrical cables, but HVDC is already pretty efficient. Would maglev applications suddenly become more widespread? Anything involving magnets (MRI, NMR) could maybe gain resolution? Are there revolutionary applications just waiting for this material?

Not at all to be dismissive, but I am genuinely curious what opportunities this could create.

View on HN

The first and easiest application would be "high Q" radio-frequency components such as antennas. I asked a similar question a few years ago a bunch of electrical engineers were basically salivating at the possibility.

The next one would be as thin films to replace copper or aluminium power delivery wires in silicon chips. This would make them run significantly cooler. While copper is a good conductor, very thin copper wires transferring multiple amps(!) is a recipe for a space heater. Even a superconductor that cost $1K per gram would be worthwhile in this application, let alone $1K per kilogram!

Next up would likely be electrical motors. Higher magnetic field strengths and lower heating losses would allow some really tiny but crazy powerful motors.

A lot would depend on the critical current and critical field that such hypothetical superconductor could have. If it's too low, its usage where large currents are needed, like motors or chips will be small. As example, look into the generation of high magnetic field. In a typical physics lab or MRI machine, the current state of the art is by superconducting coils, but for very large magnetic fields (several tens of teslas) you need to fallback on copper, since there are not known practical superconductor able to sustain such fields without losing their superconductivity

>there are not known practical superconductor able to sustain such fields without losing their superconductivity

I don't think this is still correct. Mass production of REBCO has ramped up in the past few years which is capable of 250 T internal fields and is being sold on tapes with 100,000+ A/cm^2. We're currently at the "hundreds of miles of tape per year" stage.

I think we have to be approaching the "thousands of miles per year" stage. The TF magnet demoed at CFS/MIT had 166 miles of tape and "contained close to 10% of the world’s supply of commercially available HTS" at the time. SPARC will have 18 of those magnets.

While HTS magnets may be "state of the art" they are still substantially more expensive than LTS magnets and it will be a few years before they start showing up in MRI machines.

https://www.psfc.mit.edu/sparc/hts-magnet

> Let's say there was a superconducting material that operated at atmospheric pressure and 25C. While waving the magic wand, let us also pretend it costs $1000/kg.

That depends, because you've under-specified it. Critical magnetic field and critical current density at the very least, probably also fragility on more than one axis.

Let's say it's got "enough" strength and critical magnetic field, and a density of 5 so that's $5000/litre. At a critical current of 1A/cm^2, fairly useless. At 10,000A/cm^2 that can make some neat medical equipment that no longer needs coolant that condenses the oxygen out of the air if it leaks.

It's been a while since I read about the idea — and this might have been an amateur blog rather than a serious claim — but I think 100,000A/cm^2 for reasonable density is enough to fly by pushing against earth's magnetic field?

Magnetic field of earth is ~50µT, a superconducting loop with diameter 1m (circumference ~3,14m) with 100kA will feel a force of ~15N, with 1cm² it has a volume of 0,314 liters, in aluminium thats about 0,85kg or 8,5N, in copper thats about 3kg or 30N. So with a material density of 5kg/dm³ or less it could be possible to fly. Changing the current in the superconducting coil to control lift and propulsion could be tricky though.

> it could be possible to fly

A bit of an understatement. It could be possible to fly like helicopters, but all the way into lower space, reach an orbit, stop again, go for another one, etc.

A current loop in a homogeneous magnetic field experiences a torque, but no force.

What about two loops attached rigidly to each other, with currents running in opposite directions? (I suppose that for practical stability a larger ring of coils would be needed.)

They would experience neither force, nor torque.

The resolution of MRI/NMR is more limited by pulse sequence design and magnetic gradient control. The cryogenics built into the scanner don't contribute meaningfully to those. Instead, you'd gain a huge amount of portability and/or a huge reduction in cost/complexity.

You can't spell cryogenics without "cry", and being able to sustain a large field without needing cryocoolers or a wet system would lead to impossible-to-imagine gains. Medical/industrial imaging, maglev (for sure), even fundamental physics would hugely benefit from the tech.

What about magnetic confinement fusion? My understanding was that tokamaks use superconductors to produce their magnetic fields. Does a low temperature superconductor help with that?

For NMR the resolution depends directly on the strength of the magnetic field, so if those superconductors allow higher magnetic fields they also increase resolution. See e.g. the new Bruker 1.2 GHz NMR spectrometer which uses some high-temperature superconductors while the spectrometers with conventional superconductors could not get higher than 1 GHz.

There are other effects at higher fields that are not necessarily positive, so just cranking up the field has limits. But in general higher fields provide more resolution and sensitivity in NMR, and it's quite a big increase.

Oh, that's an important correction! If this magical superconductor has a higher critical current density than current tech, then a larger field would be possible.

Also, the new Bruker systems run at 28.2 T, which is a mind-bogglingly large field. This badly structured headline made got a real double take out of me too: https://www.bruker.com/en/news-and-events/news/2022/bruker-a...

3GHz is a lot

No more gearboxes -- you can have direct drive motors controlling everything. This would revolutionize robotics, compared to current systems that require a large gear ratio between motor and joint, adding large amounts of inertia.

Can you elaborate on this? How does a superconductor help here?

The magnetic field of a coil depends on the current, and the number of turns. If you have a sufficiently thin superconductor, and keep below the critical current density, you can ramp up the number of turns until you reach the physical space limitations.

When you're working with a copper conductor, the thinner the wire, the higher the resistance, and you trade off turns for conductance (1/resistance).

Thus, a superconductor offers the ability to put many more turns, or much higher current, into a given volume.

You're limited with steel cores as to the maximum magnetic flux, but you could get the same flux with much lower power levels using superconducting coils. Alternatively, you could redesign your motor or actuator to be a coreless design, and you can then use much, much higher magnetic fields (up to the critical field of the superconductor).

In a typical brushless motor, the torque is proportional to the magnetic field from the rotor times the current in the stator (non-rotating part). The best conventional magnets and conductors (neodymium-iron-boron and copper) can only generate a certain amount of torque without overheating due to the I^2R loss in the copper. This isn't nearly enough for most applications from driving vehicle wheels to robot joints, so motors need a step-down gearbox. For something like a walking robot knee joint, it needs around 15:1 torque multiplication. This makes it move "robotically", and not be able to absorb the impacts of running and jumping.

A superconductor can handle much larger amounts of current than copper, and maybe also generate a higher magnetic field in the rotor, so you could get much higher torques directly from the motor.

Superconductors still aren't ideal -- they stop being super in really high magnetic fields, perhaps because the electrons are moving in curves rather than straight and the complicated quantum effect that causes superconductivity doesn't work. But plausible superconductors could give great results.

You could build coreless transformers that worked at 60 Hz, with no losses.

You could build electric motors without loss.

You could fill the Tevatron tunnel at Fermilab with new magnets and use it as either an accelerator, or an energy storage facility, depending on the time of day and demand.

I'd make a wedding ring out of it, and be able to feel magnetic fields.

I'd make a SQUID with it, and experiment with the |A| field (electromagnetic vector potential).

> I'd make a wedding ring out of it, and be able to feel magnetic fields.

What about biohacking? Imagine those little magnets that people like to implant below their fingertips—would a superconductor be especially cool to have there?

Maglev applications strongly depend on a sufficiently high critical field. Unfortunately, high temperature superconductors have serious issues with it.

https://en.wikipedia.org/wiki/Superconducting_magnetic_energ...

Wouldn't we be able to move power anywhere in the world losslessly? You could generate power in California meant to be used in New York for instance.

Existing HVDC technology can easily span the US with negligible losses (less than 10%).

10% is not negligible and it would be way more across continents around the globe.

But if one could build indeed a global lossless energy grid, than this would mean solar power for everyone all the time, as somewhere on the planet, the sun is always shining.

I think it is negligible. In particular, PV is much more than 10% cheaper than many other options, and this would be a replacement for whatever the round-trip cost is even in a zero-$ storage system.

My default number for HVDC is 5%/1000km, but I think that's out of date now; at that level, antipodal power would be 1-(0.95^20) = 64% loss, and while that's certainly something I'd avoid if possible, it's still good enough to make PV beat the TCO of batteries or fossil fuels or nuclear… if you ignore the TCO of installing and maintaining the cable, geopolitics, and all the other reasons we've not done that yet.

(I have no idea what the TCO of undersea HVDC cables is).

Here's one data point (330km, 700MW, 880mln EUR):

https://balticwind.eu/higher-cost-and-potential-delay-for-su...

That being said the Baltic sea is shallow, averaging less than 50m in that area.

As I read that, that's purchase price; it makes a huge difference if that lasts 10 years or 100 between replacement.

https://en.wikipedia.org/wiki/SQUID

Ultrasensitive low cost magnetometers could usher in a wave of low cost MRIs. The burden would shift from exotic hardware and power electronics to signal processing and computational reconstruction, which scales great.

Hoverboards, Railguns and Quantum computers

Now I want to read a novel with this title. Will see if ChatGPT can write one for me

It looks like it helps a lot with building long high speed data cables, depending on how you isolate the wires.

In a world where optical fiber is used for long distances, is this still relevant? For data, not for electricity.

It matters to me because it's the difference between $5 of fancy metal and $500 of transceivers to make a 10 foot thunderbolt cord.

I mean "long" in a somewhat relative sense because at tens of gigahertz your signal integrity goes to garbage after several inches, but also I found a paper talking about fractions of a decibel per kilometer if you completely isolate the wire.

Nuclear fusion strongly depends on superconductors.

Reply on news.ycombinator.com