Electric Propulsion Magnets Ready for Space Tests | Hacker News Reader

Electric Propulsion Magnets Ready for Space Tests

spectrum.ieee.org

106 pointssohkamyung67 comments

Rocket equation applies to more than just chemical fuel. The article quoted 1T for about a watt of input, which is ludicrously efficient. So the higher field means more thrust. Less input power means less power generation hardware mass and less cooling hardware mass, which will compound with the higher thrust for a very favorable specific impulse. Very exciting, and I wish them well!

Also, when did the Standard Breakfast Equivalent for torus become a bagel instead of the established donut :-)

Woke is ruining everything!

Your comment has just made me realise that when people say “woke” to mean “anything I don’t like” the subtext is “I don’t want to face up to the responsibility and effort required to participate in a decent society. I just want to do whatever I like without consequences.”

Thanks for providing the trigger for my insight.

I’m glad my stupid joke served some purpose.

Sugar tax.

But not a calorie tax.

Totally agreed. The same logic applies to really all forms of transportation. Mass begets mass.

Mass baguettes mass!

Experimentally baguettes are bad for superconductor operations: https://www.theregister.com/2009/11/05/lhc_bread_bomb_dump_i...

So is this just a more efficient form of ion propulsion? Or does it have enough thrust to potentially supplant chemical rockets for some lower-thrust use cases?

> Rocket equation applies to more than just chemical fuel

But it doesn’t apply when you’re collecting energy in situ. The article proposes solar panels. Rocket/wagon equation doesn’t apply.

What rocket equation cares is reaction mass not energy: ve * ln(m0/mf) where ve is exhaust velocity, m0 is initial total mass, which is dry mass + propellant mass, and mf is dry mass

> What rocket equation cares is reaction mass not energy: ve

Break out exhaust velocity.

Surely a n00b question, but: why superconducting electromagnets?

I get that means zero electrical resistance and stronger magnetic field at rest, but ultimately the engine is doing work accelerating ions, which will be "felt" as resistance in the electrical system. The total energy must be conserved. I would imagine (?) that work done accelerating the ions is far greater than the losses to heat in the magnetic coils.

Or maybe that's not right?

Less energy use, lower mass? I know superconducting magnets are a goal for both aircraft propulsion systems and use in wind turbine generators, just for the mass reduction.

Generating a strong field requires a lot of current. If you have any resistance at all, that's a lot of heat dissipation, which will require cooling.

Here's a link to the actual paper [1]. According to the paper, the magnetic part of the thruster consumes between 10 and 100 kW if it uses regular magnets, and only 0.2 kW with these superconducting magnets. Most spacecraft use solar panels, and they claim the largest satellites produce 30 kW of power. It follows that 10 kW is an unacceptable level of power consumption, while 0.2 kW is ok. They admit that nuclear powered spacecraft don't necessarily need this new technology, because 10 kW might be negligible for them.

[1] https://www.sciencedirect.com/science/article/pii/S277283072...

Wouldn’t it still be better to use more efficient engines? could they not just have more of them?

You’d get more thrust, but lower delta-v for the same total mass, because you’re replacing propellant mass with fixed mass.

That is an economic question. What does efficiency costs vs what does providing more power cost?

There are several ways a spacecraft engine can be efficient. The fundamental thing is the tyranny of the rocket equation: in order to get faster you need fuel, and then you need fuel to push that fuel, into a compounding way. The best way to address this tyranny of the rocket equation is to have an engine that expels something at a very high speed. The best chemical rockets push hot gas at about 4.5 km/s. The engines discussed here are "magnetoplasmadynamic thrusters" [1]. They push plasma with a velocity in excess of 15 km/s. Some studies suggest even 100 km/s. They are basically the most efficient rocket engines our current technology can build.

The next type of efficiency is how much of the source energy is converted into the energy of the plasma jet. Is it 50%, 90%, 99%. That I don't know, and obviously you want as much as possible.

But this project is addressing a different problem. In order to create this plasma jet, you need a very strong magnetic field to contain the plasma while it is being accelerated. The field itself does not provide the propulsion, it is there only for containment. You don't want to expend a lot of energy maintaining this field. Superconducting magnets are much better than regular electromagnets, so it's no surprise that containing this plasma with superconducting magnets requires a fraction of the power of regular magnets. But someone still needed to do the actual research. What works on paper doesn't translate always in things that work in real life. These guys did just that, and now are ready to send a technology demonstrator in space.

[1] https://en.wikipedia.org/wiki/Magnetoplasmadynamic_thruster

Power consumption may not be the only concern but also dissipating the heat. You don't have air or water around you to dump the heat into so you're limited by what you can radiate (is my understanding). Trying to radiate out 200W of heat vs 10-100kW is a dramatically different challenge.

Fill a surrounding container with a volumetric shit ton of nitrogen gas (perhaps with stepped pressure gradients) and you might be able to radiate a significantly larger amount by increasing your surface area without drastically increasing your weight (or use nuclear power to mitigate the weight concerns).

But yes, today’s design space generally prefers as low a power envelope as possible to not have to worry about dissipation.

Wouldn't the limiting factor be the outer layer? I'm not sure what layering would give you, or what adding a container would give you. For short bursts this may make a difference but the steady state means the system as a whole needs to radiate out that energy somewhere. More mass gives some kind of ballast I guess but kilowatts of energy is just a lot to get rid of.

It's not about ballast. Black body radiation, which I'm sure you know is the only way heat is dissipated in a vacuum (unless you're the Sun & even then it's still the primary way), is defined as:

P = ε * σ * A * T^4

That's the constant, the emissivity of the material, the surface area A, and temperature.

We want to increase the temperature differential above what the natural black body radiation can dissipate for a given device (let's take a space probe) and assume we can't change the material. The only thing we can play with is surface area.

The point of the nitrogen gas is to leverage normal convection / conduction to dissipate into a larger gaseous volume which then naturally also has a larger surface area. This would then give you more surface area to dissipate across. The reason I'm thinking the layering with different pressure gradients might be useful is that it reduces the overall weight because you need less overall gas for a given dissipation profile (i.e. you need high density beside your heat source to transfer energy away quickly but less density further away because there's a larger surface area growing with the square of the distance away from the heat source).

[1] https://en.wikipedia.org/wiki/Stefan%E2%80%93Boltzmann_law

Convection depends on buoyancy and hence gravity, and conduction in nitrogen gas is extremely low, so I don't think your idea would work. You need a conventional radiator with forced circulation.

Fractal branching structures (trees), which maximise radiating surface area, don't make good pressurised vessels.

Inflatable envelopes in space are challenging (micrometeoroid damage). The envelope, and thus the gas, need to be hot, lest T^4 swamp R^2. It's feasible to cool a reactor to low-pressure gas (Mars atmosphere [1]) using forced convection, but that's cold gas and IIUC the heat exchanger is comparable to the size and mass budgets of a radiator. So attempting a steelman of a set of forced-convection radiators on poles, hosing hot gas against the envelope... hmm. Big picture, this attempts a low-mass heat transfer construct, using a large-scale low-pressure gas flow, where the cold-side surface is nearly as hot as the hot-side, and is extremely mass constrained (thus flat-ish). Seems hard? Fun thought experiment - thanks!

[1] Colgan Mass Optimization [...] Forced-Convection Heat Exchanger for Mars Surface Waste Heat Rejection 2023 https://asset.library.wisc.edu/1711.dl/FYJCG5YME7O4X82/R/fil...

VASIMR is one of the few test articles with flight hours of this category (or maybe a sibling category), and when you factor into account the thruster and the power supply, its thrust to weight ratio for the VX-200 prototype was unworkably low. Not "too low for launch", I mean "too low for a steady burn of months to attain meaningful progress towards interplanetary or orbital objectives".

You have to simultaneously solve for power to weight ratio issues on solar panels, power supply, and the actual thruster. The best solutions right now for most purposes are relatively low Isp (low propellant efficiency) Hall thrusters.

In space, heat dissipation is remarkably harder than it is on Earth, so getting to, for example, 97% thermal efficiency rather than 94% would be a big deal in terms of the mass of the heat dissipation system, and I think this and reducing the mass of the actual electromagnet is the advance here, if there is any.

One of the biggest things that always bothers me in sci-fi, even harder sci-fi like The Expanse, is the missing heat sinks. All those ships would need massive radiators or they’d melt.

The starship in Avatar got that right. Rest of the film is comic book level and derivative but the ship is good.

It's worse than you think. Here's a really interesting article on what it would take to get Epstein drive performance:

https://toughsf.blogspot.com/2019/10/the-expanses-epstein-dr...

This thing would produce more power than all of civilization does today. No radiator could handle it. They use laser fusion, exploding D-He3 fuel pellets well behind the ship, with a giant magnetic nozzle. A tungsten heat shield protects the ship, and reaches an equilibrium temperature without active cooling.

In the Expanse though The Drive is always on, which means the ship is tossing propellant mass (efficiently) out the back of it - which means it can dissipate most of it's heat that way.

The Drive in the Expanse is _nonsensically_ powerful; Turning it on at mass fractions that seem compatible with their ship designs would be akin to a nuclear exchange for anything within thousands of kilometers of the rear end.

Talking about thermal and propulsive efficiency for a rocket is a little tricky because there are multiple useful metrics, but in this context "waste heat" is anything that doesn't get tossed out the back.

The magnetic field here is acting as a nozzle. The circuit driving the field is not doing work on the ions, and any power loss is just parasitic. It's generally not worth it for small thrusters, because the extra efficiency gained by having a magnetic nozzle is offset by the parasitic mass and power cost (here they eliminate the power cost).

Would sure be nice to have some numbers on expected thrust, weight, and ISP, in order to have some kind of context in which to understand this!

its not been tested yet, so they probably have no idea.

They for sure expect better than the ones currently in use/testing (e.g. the nstar ion thruster)

> very stringent stray magnetic field requirements of the ISS.

I'm sure that others, here are far more conversant with the tech, but it's my understanding that one of the attributes of superconducting magnets, is that the very strong field is extremely localized. It doesn't extend too far from the magnet.

I thought the strength of most(all?) fields is inversely proportional to the square of the distance from the center/source? I think this is called inverse square law.

Its also for these reasons making a very large nuke is pointless, after some distance the effects are almost 0.

Yes, that's what I would think, but I remember reading that, and thought it sounded strange. The article made it seem as if the rules were different for superconducting magnets.

Might just be a manifestation of the Inverse Square Law. Like I said, not my area of expertise.

Very large nuclear bombs are inefficient because most of the energy goes into heating up air, so not relevant to the current thread.

They're only inefficient if you want them to do something other than heat up a lot of air by a very large amount very fast. :)

OK: they're inefficient at killing people and destroying buildings and other structures relative to bombs with yields like 500 kilotons.

Pretty efficient in their mass to destruction ratio, it's only inefficient if you compare it to how much energy the fissile/fusible material could be worth in a nuclear reactor.

It's also an inefficient use of ICBMs or whatever "system" is being used to "deliver" the bomb to the target (which BTW is more expensive than the bombs themselves).

Well that heat does a lot of damage too, on top of other damaging radiation, blast wave, irradiating everything etc. which is proportional to bomb yield.

Sure the biggest ones are comparatively ineffective, thats why the focus on manufacturing was on small/medium yield one, one can produce many more and saturate defenses more easily rather than one/few MOABs.

Agreed that I know very little about this subject.

But heat itself can be imagined as a kind of field. For eg: A candle, the closer you move to the flame, every point closer to flame, the temperature is higher. The farther you move apart the temperature reduces. Same with the Sun.

Take for eg- Sound. The closer you move to the source of it, the louder it sounds, the farther you move from it, the fainter it gets.

Now you could call this a wave. Like the wavelength stretches with distance, also decreasing frequency and amplitude. Another way of looking at this a field. Imagine 3D space with points, like a lattice. Each point has a value, and the values get denser to the center, and move farther apart as move away from center.

If Im not wrong a lot of physical phenomenon exhibit this behaviour. Including thing like light, sound, temperature, gravitation etc.

While you can make a beam of light, I doubt you can make a beam of magnetic field(ray?) or gravity for that matter?

Heat is only a field in radiation because it’s literally being carried by electromagnetic waves. Heat is also conduction and convection - in those cases it’s decidedly not a field.

The heat and wave equations have an important difference:

    heat: ∂u/∂t = k ∂²u/∂x²

    wave: ∂²u/∂t² = c² ∂²u/∂x²

Note the extra time derivative on the wave equation. The heat equation is dissipative, moving towards the average value, spreading energy out over time. The wave equation, on the other hand, allows for solutions that oscillate and maintain their shape while they travel.

This means that waves such as electromagnetic (light) or gravitational can in theory be focused and beamed significant distances without much loss. A plane wave solution is lossless, but requires an infinite emitter.

Finally, note that this does not apply to static fields, such as magnetic, electric, or gravitational fields, which if not oscillating will still spread out and dissipate over distance.

So the second equation does not cover my induction cooktop?

There are many efficiency metrics. For instance, how much energy went into constructing and deploying the bomb vs. how much destruction it caused. I think large nuclear bombs would look quite ok on this metric.

People have known how to make very large bombs for a long time (e.g., in 1961, the USSR tested a bomb with a yield of 50–58 megatons) so it says something that for decades, no military has had any bombs larger than about 1.3 megatons.

Magnetic fields produced by dipoles have an inverse cube dependence, so dissipate much faster than gravity or other fields.

If we ever find a monopole, the field will be inversely proportional to the square of the distance.

Bug normal magnets and superconductor magnets have canceling magnetic charge, the same amount of S and N, so after a few math tricks [1] the field will be inversely proportional to the cube of the distance.

[1] The filed from N decays as 1/r^2 and the field form S as -1/r^2 but they are indifferent places (the r is not the same on both), so the difference is almost like 1/r^3 more details in https://en.wikipedia.org/wiki/Multipole_expansion

Thanks for the link.

Sometimes I wish Xanadu had actually happened. The moment I opened the article I saw something like |R^3. Honestly speaking I don't really know what |R means. And R keeps showing up all over the article.

I will dig more into this using ChatGPT with time.

But I could understand very little from reading the article.

Blackboard bold R with a superscript 3? That's just space, like regular space that we inhabit. The 3 is the dimensions, the R is the real numbers, and this distinguishes it from other wacky kinds of space that mathematicians like to try things out in.

Edit: we also have in there "the coefficients of the multipole expansion can be written as functions of the distance to the origin, r", so it sounds like r in italics is the origin - but that's usually (0, 0, 0), so maybe it's a distance to it. Anyway r in bold is probably a vector, vectors are usually in bold. And you're quite right about Xanadu, I wonder if there's some way to overhaul Wikipedia to put a key to notation in the sidebar of all pages with equations in.

>>and this distinguishes it from other wacky kinds of space that mathematicians like to try things out in.

What other space apart from 3 dimensional space exists? And what does that even mean?

Math doesn't have the same limitations on dimensionality that our brains and the space that we interact with do. A line becomes a square becomes a cube becomes a hypercube, and so forth.

Well, the nicest ones are two dimensions and one dimension, which sort of (not really) exist as subsets of 3D. So a single variable is one-dimensional. Then you could have a gradient, say:

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

And you could have a gradient (of colors if you like, or salinity or force of whatever) along a line, one dimension, or on a plane, two dimensions, or in space, three dimensions. By extension, just adding components to vectors and iterating over the components when you do stuff with them, you can apply the same logic about higher dimensions. This makes people thin and unhealthy and serves no purpose, but seems to have a compulsive appeal for a certain kind of mind.

You can't always reliably extend some idea to different numbers of dimensions, because the interactions between components can get complicated, sometimes a thing that's valid for certain numbers of dimensions is not in fact valid for other numbers of dimensions, like maybe you have to have a odd number of dimensions to prevent positives and negatives cancelling out or something. Like many people, I knew this stuff for about a month once but forgot the specifics. But a tesseract is kind of fun to look at.

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

This is like the next step in the sequence dot, line, square, cube. The animation is a "projection into 3 dimensions", like how the shadow of a cube on a piece of paper is in two dimensions. It can't be shown in four dimensions, due to four dimensions being complete fiction.

(It's squirming like that because it's rotating, but the axis it's rotating around is at some 4D angle.)

> What other space apart from 3 dimensional space exists? And what does that even mean?

Aside from the one and two-dimensional spaces mentioned in a sibling comment (ℝ¹ and ℝ²), there's also spaces with more interesting geometry.

For example, S² is the two-sphere, more conventionally known as "the surface of a ball." On the sphere, the equivalent of straight lines are the great circles. S¹ would be the one-sphere (the circumference of a circle), and S³ would be an interesting, periodic three-dimensional space – kind of like what our universe would be if it were closed rather than (probably) open.

Here I will shoehorn in a mention of Hyperrogue, which is almost related.

https://www.roguetemple.com/z/hyper/

The field generated by superconducting magnets have the same properties as any field generated by any other kind of magnet.

The magnet's geometry and intensity are what define the field, not the material.

How long will it be before this new space propulsion capability can be scaled in order to put the ISS on the moon instead of in the ocean?

Long after it already is. The ISS is aging, and there was every intention of retiring it even before its principal sponsors started a proxy war.

There's certainly zero chance that it could be on the moon. It wasn't designed to survive on a surface. It would not be able to support itself.

If we want something in orbit around the moon, it will still be far cheaper to build a new thing designed for that purpose.

It does seem a shame that it can't be recovered and donated to museums or preserved in some way. I assume that it would be prohibitively expensive/complicated to try to do so, but it's a huge part of the history of space research, and it's a bit of a bummer to just throw it away.

I mean, if Starship works well, it could theoretically retrieve the ISS in a piecemeal fashion.

Orbital refuelling of non-satellite spacecraft at greater than IDK a few km has not been demonstrated.

Rendezvouzing and pushing or pulling a 900K lb (350K kg) object in orbit has never been demonstrated.

In-orbit outer hull repair on a vessel with occupants has also never been demonstrated?

Are there NEO avoidance plans that do not involve fracturing the object into orbital debris on approach?

Reply on news.ycombinator.com