Fusion aerospace propulsion systems for the near future

[Xforce]

Let’s ditch chemical fuel (even like scramjet, aero spike and SABRE) and venture to the future, for only thus we can become a multiplanetary species, and establish a true interplanetary/interstellar transit system.

For atmospheric ascent:
1.Fusion reactor (thermal power)+Thermal ramjet (inspired by KSP interstellar): use the heat of a fusion reactor (can be an easy D-T fusion) to heat up atmosphere gas to a high temperature, then they expand in volume, creating a thrust similar to the rocket engines we have now. The higher the speed and denser the atmosphere, the more thrust it produces. I think it have a relatively low Isp, but since the air is not carried onboard, efficiency don’t matter. DT fusion gets 3.401*10^14J per kilo, that’s enough to get you outta the Kerman line. And I’m sure it works at any planet with an atmosphere, Venus, Titan, and it can scoop up fusion fuel at the gas giants. But the nozzle can be easily overheated.
2. Fusion reactor (electricity) +MHD: while I’m don’t know much about this method, but I really like it as an eco-friendly electricity propulsion method. Works pretty well from subsonic speeds to hypersonic, high altitudes. On the official definition it’s the “manipulation of fluid by electromagnetic forces” and maybe it involves ionising air then accelerate it to hypersonic speeds/orbital velocities/fraction of c? Need some help for this one.
For space flight:
Inertial fusion engines: while currently experimental prototypes only works in vacuum, and have a limited Isp of around 7000s (fusion is is millions times more energy dense, but now only around 20 times the efficiency) we still expect it to become it dominating space travel. When using Deuterium-Helium 3 the exhaust (as all the products are charged particles, it’s easy to direct /vector thrust , in theory can reach up to 9.8% c, or convert into 3 million seconds of Isp (maybe?) by the time we can mass production those engines, interplanetary trips will be quite easy as you accelerate half the trip at 1g, turn retrograde and decelerate half the trip again at 1g

What other propulsion systems can be available in the near future (around 50 years)? And how does MHD works?

 

[member 656954]

Fun to think about but it's unlikely we'll have fusion powered transport (or anything?) of this kind, even in 50 years.

One of our best bets at fusion is ITER, which is not small at 15.5m high and 23,000t in weight. I get that it's experimental and I'd expect - if it works - that fusion reactors will be made smaller as time goes on, but we'll likely burn 20 - 25 years of your 50 years just to get ITER ignited and proving the fusion concept.

As with fission reactors, the economics will favor large reactors over small reactors, but given the complexity and precision engineering needed, they won't be inexpensive for many decades after ITER ignites (if it ignites). So the pressure to create a 'small' fusion reactor suitable for a spacecraft will not emerge until close to the end of your 50 year window. By then, I'd be expecting solar powered craft utilising 90% yield PV panels with high energy density batteries to be in use so the cost effectiveness of fusion powered craft may never eventuate.

But that's all speculation and I'm not even sure it belongs in this forum

 

[Xforce]

Fun to think about but it's unlikely we'll have fusion powered transport (or anything?) of this kind, even in 50 years.

One of our best bets at fusion is ITER, which is not small at 15.5m high and 23,000t in weight. I get that it's experimental and I'd expect - if it works - that fusion reactors will be made smaller as time goes on, but we'll likely burn 20 - 25 years of your 50 years just to get ITER ignited and proving the fusion concept.

As with fission reactors, the economics will favor large reactors over small reactors, but given the complexity and precision engineering needed, they won't be inexpensive for many decades after ITER ignites (if it ignites). So the pressure to create a 'small' fusion reactor suitable for a spacecraft will not emerge until close to the end of your 50 year window. By then, I'd be expecting solar powered craft utilising 90% yield PV panels with high energy density batteries to be in use so the cost effectiveness of fusion powered craft may never eventuate.

But that's all speculation and I'm not even sure it belongs in this forum

However, fission reactors are energy dense, but not quite as dense as fusion. In engines used on ships like aircraft carriers or nuclear submarines, fission engines have the least power to mass ratio (due to heavy radiation shielding), something that shouldn’t be flying. Plus, if you use a nuclear fission thermal ramjet, you will get huge amount of radiation going through the air, pretty like the Chernobyl power plant, even when it’s working normal. And when it crashes (all spacecraft get into accidents sometimes) it will be a complete☢disaster. In this case, I would rather prefer using chemical engines.
Current fusion reactors are bulky (like ITER) just like the first computers being made. But since all the products produced in D-He3 are charged particles, we can transform their kinetic energy to electricity, or just redirect them using magnets. This doesn’t require any shielding. (The ITER use DT fusion, so neutron shielding is necessary) and even when they are damaged, the reaction simply stops. All the products it have is harmless helium...
And you want 90% efficiency photovoltaic solar panels? Well, that’s hard. The mediocre panels we buy or used on satellites gives an efficiency of 20-30%, cutting edge PV in development that diverts light spectrum gives an efficiency of around 40%-50%, and still highly unreliable and expensive. Which are FAR away from 90%. To achieve that I think you have to beam lasers from the ground or from satellites, having a single wavelength will be much easier to deal with.
Btw, even the most energy dense battery we possibly have is still far away from fusion... lithium batteries (our best) yields 6.12 MJ/kilo, while an average D-T fusion at 30% efficiency will give off 10^8MJ per kilo...

 

[jackwhirl]

Did you want to limit this to fusion powered options?

I find these interesting, and you could power them with fusion, I suppose:
Atmosphere-breathing electric propulsion
Beam-powered propulsion

 

[russ_watters]

IMO, air-breathing engines are a non-starter because of their extremely limited domain of applicability. Rockets just aren't in enough air for long enough for them to be useful.

 

[jackwhirl]

IMO, air-breathing engines are a non-starter because of their extremely limited domain of applicability. Rockets just aren't in enough air for long enough for them to be useful.

I'm with you on this for rockets, but not satellites.
Starlink is already taking a step to save costs by using krypton instead of xenon in their satellites. Switch to air-breathing and those things could hang out in VLEO indefinitely.

 

[Xforce]

Did you want to limit this to fusion powered options?

I find these interesting, and you could power them with fusion, I suppose:
Atmosphere-breathing electric propulsion
Beam-powered propulsion

Both thermal ramjets and MHD are air-breathing engines. And MHD is the air-breathing electric propulsion (my favorite in+atmosphere propulsion system!) additional to that, maybe you can ionize air to make plasma, then accelerate it to extremely high velocities, like a coil gun?

IMO, air-breathing engines are a non-starter because of their extremely limited domain of applicability. Rockets just aren't in enough air for long enough for them to be useful.

It make sense to take advantage of the atmosphere on the way ascent... and when into mass production, these air-breathing electric propulsion engines will have a variety of use, small as a MHD jet pack, and large as a hypersonic airliner.
Plus, vacuum optimized engines like inertial fusion will not work in atmospheres of planets like Earth and Venus, I think it might even have trouble working on mars...
Therefore, having a hybrid system makes sense, as it offers maximum efficiency. A single propulsion system works in all regions may be lightweight and cheap, like chemical rockets, but their efficiency just suck at both. At the current technology, most rockets have a specific impulse of 300-400 seconds, much inferior to a atmosphere optimized turbofan, even a scramjet far surpasses that; and in space, it’s no match for ion engines in terms of efficiency. It may have a lot of thrust but the efficiency...ugh, we need 90% weight of rockets just for fuel.

 

[member 656954]

And you want 90% efficiency photovoltaic solar panels? Well, that’s hard.

Yeah, but you've put this speculation 50 years out and we're already seeing multi-layer PV that harvests different wavelengths so I think it's fair to push the envelope in a tech we have when you're introducing a tech we don't have some of which is based off a computer game

(And I've a similar thought on battery energy density!)

Aside from that, perhaps a useful question to start with is "What is the smallest likely size we can make a fusion reactor to power a plane?"

ITER and JET are pretty big, but some of the non-Government players working on fusion have smaller designs:

• Tokamak Energy - looks about the size of a bathysphere
• AGNI Fusion - hard to tell, it's still conceptual, but it looks pretty compact, perhaps 4m long
• General Fusion - looks to be about the size of a couple of jet engines
• Commonwealth Fusion Systems - it's reasonable sized device, definitely bigger than jet engines but nowhere the size of ITER

So on that basis, assuming any of them can get fusion working, you may have an energy source that is compact enough to fit into a plane.

1.Fusion reactor (thermal power)+Thermal ramjet

There was that nuclear fission bomber prototype in the 50's, so that suggests the concept is feasible for flight - if not a ramjet - but how exactly you conduct the heat transfer would need to be worked out. Shintaro Ishiyama, Yasushi Muto, Yasuyoshi Kato, Satoshi Nishio, Takumi Hayashi, and Yasunobu Nomoto proposed supercritical CO2 for electricity generation off a fusion reactor which may work well for this use case, but I don't have the math for whether you can transfer enough energy to drive a plane to ramjet speeds.

Note though that the energy density of fusion reactions in gas is less than fission reactions in solid fuel, so fusion reactors need to be bigger than a comparable fission reactor to generate the same power (think I got that right). So my idea of a 'small enough' reactor based on those current ideas above may not hold.

The other issue is physical movement, because the magnetic containment method needs to work in an accelerated reference frame. None of the current designs are trying to solve that problem, and it may that keeping the flickering fusion flame alive while in motion is beyond our engineering, even in 2070.

2. Fusion reactor (electricity) +MHD

What, like the ionic wind method, like what MIT built last year? Now I'll admit to using 'ionics' for a plane in my sci-fi books, though including a fusion reactor to power the beast didn't occur, even to me! Whether even a fusion reactor can create sufficient electricity for an ionic flow that generates lift enough for a large aircraft would need to be worked out. You could probably conclude 'yes it can' with enough assumptions, but it would be entirely speculative until we - if we - get fusion working.

And how does MHD works?

Well, the Wiki article for that seems a good place to start if you've not seen it before, and if you have, there are a lot of links to reference literature to work through.

 

[mfb]

IMO, air-breathing engines are a non-starter because of their extremely limited domain of applicability. Rockets just aren't in enough air for long enough for them to be useful.

Falcon Heavy burns ~500 tonnes of fuel before it reaches 10 km altitude, about 1/3 of its total mass. By the time it reaches 25 km (which still allows aircraft to fly) it burned half of its mass. The numbers for other rockets should be comparable (not considering pure solid fuel rockets). And that rocket is not using air - an air-breathing rocket has a different flight profile.
Single-stage rockets are an interesting concept because of their simplicity, but they need every little bit of performance boost they can get on Earth. That's why SABRE is planned to use the atmosphere as long as it can.

----------

I don't expect to see fusion reactors on spacecraft within the next 50 years - but who knows. They get easier to build if they are larger, but putting an ITER-sized fusion reactor on a spacecraft doesn't sound useful. Let's look at numbers, for a pure rocket first: We want 600 seconds I_sp, we need enough thrust to take off, and let's give the fusion reactor x=15% of the total rocket mass (5% more are structural mass and payload, leaving 80% fuel). This means we need a power density of g²*I_sp/x = 400 kW/kg. Better add 50% if you actually want to launch instead of hovering. This is a ridiculous power density. ITER is supposed to produce 500 MW thermal power - you would have to shrink it to about 1 tonne!
How can we lower this? With a lower I_sp we need more fuel and x goes down, with a higher I_sp x can go up but now I_sp is larger. My numbers are maybe not the exact optimum but they are not far away from it either. Doesn't fly.

Air-breathing you can (a) increase the mass fraction of the fusion reactor, (b) lower the I_sp because you have more reaction mass and (c) lower the thrust to weight ratio with wings. All these things make it much more realistic. Reach hypersonic speed in the atmosphere, afterwards use your own propellant to go to orbit.

----------

In space you can get away with much lower thrust. Shielding is also less of an issue and a launch failure doesn't make your launch site radioactive. I still don't expect fusion reactors any time soon there. Fission reactors are so much easier and they have more than enough energy density for interplanetary flights. The limits are not coming from the amount of fuel.

ITER and JET are pretty big, but some of the non-Government players working on fusion have smaller designs.

None of which have ever produced any relevant fusion power.

 

[member 656954]

None of which have ever produced any relevant fusion power.

Nope, but really, neither has ITER and even JET is light on 'power'. Given that @Xforce put the scenario 50-years out, if we get fusion working, some of these designs are equally likely to also be working. But I'm equally skeptical, fusion is a really hard engineering problem and it's probable that even if ITER, JET, China EAST, or one of those new firms fires up a sustained plasma, it may not be economic in the face of the RE juggernaut to warrant building many more reactors.

Fission reactors are so much easier and they have more than enough energy density for interplanetary flights.

Absolutely, and if we can construct ships in orbit, or on the Moon, or on some other moon, then load them up with enriched fuel, the reactor never even start up anywhere near Earth such that a meltdown, or containment loss, or some other failure is a local tragedy, rather than a widespread one.

 

[russ_watters]

It make sense to take advantage of the atmosphere on the way ascent...

I'm not sure if you recognize just how little time such an engine would function for a rocket. Unless the ascent profile were altered to try to maximize its use (without tearing apart the rocket), a rocket would spend maybe a minute - the second minute - of a 9 minute ascent in an envelope suitable for scramjet use. They could already be using turbojets for the first minute, in thicker air, if it were viable. I think it is just more trouble than it is worth.

 

[mfb]

Nope, but really, neither has ITER and even JET is light on 'power'.

2/3 for JET, ITER should get a factor 10. The small projects are not even close to these numbers (at least with verifiable results...).

Given that @Xforce put the scenario 50-years out, if we get fusion working, some of these designs are equally likely to also be working.

Why?
Following the same logic, if a nuclear weapon with 50 kg uranium works then a nuclear weapon with 50 grams of uranium should work as well? Clearly not. There is a reason ITER is larger than all previous machines. The volume to surface area ratio is important.

Unless the ascent profile were altered to try to maximize its use

Of course it would. See the Skylon concept. Fly and accelerate to ~2 km/s in the atmosphere, then go from there like a conventional rocket. The mass saving is not that big (especially as you have to add wings), but big enough to make an SSTO rocket useful - at least on paper.

 

[Xforce]

2/3 for JET, ITER should get a factor 10. The small projects are not even close to these numbers (at least with verifiable results...).Why?

For aerospace grade compact fusion reactors, they should use deuterium-helium 3 fusion instead of D-T where both ITER and JET uses D-T. In a D-T reaction, the thing sucks is that the 14.1 MeV out of 17.8 MeV per fusion reaction was carried by an annoying neutron... and we have nothing do do with it, we can’t manipulate nuclear force to make it’s kinetic energy any use. 80% energy wasted, that’s why most fusion reactor (other than the gargantuan ITER) outputs less energy than inputs.

Plus, we have to have heavy neutron shielding in this case, where non-charged neutrons have more penetration than gamma rays, the shielding drastically lowered the energy density. These utilizes DT fusion considers easier fusion (a promising start) , while aneutronic fusion is harder, so we need to wait more years until it coming.

By using D-He3 fusion, where it’s a bit hard to fuse, the fuel is rare (tritium was expensive as well) but at least all the energy is carried by charged particles... which means more efficiency, as we can manipulate electromagnetism, and convert their kinetic energy to electricity.

To see what technological improvements can happen in 50 years, you can compare the Flyer by Wright brothers and the SR-71 or X-15 (from propeller to TBCC, from bi-wing to lifting body, from 30 seconds in air to Mach 3 cruising), or the computer on the Apollo mission (much inferior than the calculators we use) and a modern day iPhone (small and portable).

In space you can get away with much lower thrust. Shielding is also less of an issue and a launch failure doesn't make your launch site radioactive. I still don't expect fusion reactors any time soon there. Fission reactors are so much easier and they have more than enough energy density for interplanetary flights. The limits are not coming from the amount of fuel.None of which have ever produced any relevant fusion power.

Fission thermal engines like NERVA was inferior than inertial fusion engines or even ion engines we have have around 4 times the efficiency than that.
Even a long time before using fusion powered SSTO, we could have fusion powered spaceships, constructed in orbit (of course carried by advanced chemical rockets like Starship) as they required much less energy density than SSTO. They can still get us to Mars in a week or so.

 

[member 656954]

Why?
Following the same logic, if a nuclear weapon with 50 kg uranium works then a nuclear weapon with 50 grams of uranium should work as well? Clearly not. There is a reason ITER is larger than all previous machines. The volume to surface area ratio is important.

So are you saying that the physicists working on those machines are misguided? I'd expect they have reason to believe their designs will ignite and generate power, and are not making such an obvious flaw that you can pick it from a PF post that references them in passing. Besides which, we're not talking 50kg to 50g, you're making a false equivalence.

As for ignition, ITER hasn't fired up either, but the point is that it's not the only fusion player in town and other designs are being tested that have different characteristics. Maybe they'll work, maybe they won't. But given that ignitions have been few and fleeting, I certainly wouldn't be betting that we've nailed the "Highlander" solution this early in the game.

 

[mfb]

@Xforce: D-He fusion is much more difficult than D-T fusion. So much more difficult that it is unclear if any reactor can keep it burning even under ideal conditions. See if the concept works with D-T fusion. If the answer is "no" even there then D-He fusion has no chance.
The energy of the neutron is not wasted, it's the main output of the fusion reaction. The other 20% are used to keep the plasma burning.

D-He fusion is not aneutronic, by the way, you get D-D side reactions.

@Tghu Verd: I'm not saying they cannot work. They are high-risk projects: Small chance of success, great result if they succeed. Claiming "if the big machines work then the other approaches must work as well" is wrong.

Besides which, we're not talking 50kg to 50g, you're making a false equivalence.

I thought it was clear that these numbers were not meant to be a 1:1 relation. They were about the general concept: Not everything scales nicely.

 

[member 656954]

Claiming "if the big machines work then the other approaches must work as well" is wrong.

Apologies if you read that into my post, @mfb. I was merely noting that smaller-than-ITER designs are being planned which might fit into an aircraft (or rocket, though if we're not launching from Earth that's less of an issue). I doubt that they'll all transition to working reactors...in fact, I'll be surprised if any of them do, you need to be richer than Scrooge McDuck to fund the R&D and chances are it aint gunna work in any event because fusion is so hard to contain and sustain!

I thought it was clear that these numbers were not meant to be a 1:1 relation. They were about the general concept: Not everything scales nicely.

Sure, but I reiterate that if someone is working on a physical reactor right now that is considerably smaller than ITER and JET, then presumably they've some idea that their smaller design has a chance of working. You're suggesting these smaller designs won't work because you can't scale fusion down and that's a given. Which means either you're mistaken or physicists at the likes of Lockheed Martin, Commonwealth Fusion Systems, and General Fusion are mistaken. You can't all be right!

 

[mfb]

You're suggesting these smaller designs won't work because you can't scale fusion down and that's a given.

No, I'm saying the logic "if large reactors work then small reactors are equally likely to work" is flawed. Here is the original text I objected to:

if we get fusion working, some of these designs are equally likely to also be working.

The physicists at these projects won't tell you "if ITER works then our design must work as well".

 

[member 656954]

Ah, I see, @mfb, thanks. Yes, I should have put more words into the sentence, which was intended to convey that solving fusion by any method is very likely to lead to other working methods because it is easier to engineer from a proven design than from a theoretical design.

 

[Xforce]

@Xforce: D-He fusion is much more difficult than D-T fusion. So much more difficult that it is unclear if any reactor can keep it burning even under ideal conditions. See if the concept works with D-T fusion. If the answer is "no" even there then D-He fusion has no chance.
The energy of the neutron is not wasted, it's the main output of the fusion reaction. The other 20% are used to keep the plasma burning.

D-He fusion is not aneutronic, by the way, you get D-D side reactions.

@Tghu Verd: I'm not saying they cannot work. They are high-risk projects: Small chance of success, great result if they succeed. Claiming "if the big machines work then the other approaches must work as well" is wrong.I thought it was clear that these numbers were not meant to be a 1:1 relation. They were about the general concept: Not everything scales nicely.

D-He3 requires 2-3 times the energy to ignite the reaction (which one will we have sooner, D-He3 reactors or He3 mining facilities on the moon?). D-D reaction (which results in tritium and DT fusion) does occur, but not so common. Around 5% the total energy will be carried by neutrons produced by DT fusion (this was from another source, I didn’t calculate myself) with 95% energy carried by charged particles.
Plus, how can you make the power output of neutrons work? You get heavy radiation shielding, then use a thermoelectric generator to convert the heat to electricity? If so, then the reactor must be massive like ITER
By the way, this topic have become more nuclear physics than rocket science

 

[mfb]

D-He3 requires 2-3 times the energy to ignite the reaction (which one will we have sooner, D-He3 reactors or He3 mining facilities on the moon?).

Requires more energy and produces much less. Here is the classical plot. The peak is at ~3 times the temperature which means ~80 times the radiation losses, with a factor ~7 lower cross section. The reaction releases the same energy but all of it can be used in the plasma, you gain a factor 5. Overall you are worse off by a factor of about 100. This is not a factor 100 in power production, it is a factor 100 in the power to keep the plasma burning. Yeah... probably not, unless some revolutionary new concepts can be found that make it realistic.

Plus, how can you make the power output of neutrons work?

The neutrons hit the blanket and heat it, the heat is extracted with a liquid (water, lithium, ...) and converted to electricity. The reactor doesn't have to be massive for that.

By the way, this topic have become more nuclear physics than rocket science

Yes, that happens if all your questions are about fusion.

 

[jackwhirl]

Hey, @mfb, can we take this thread in a slightly different direction?

I'd like to explore ITER sized or larger reactors on spacecraft that are meant to be constructed and operated in space and never ascend through the atmosphere.

I'm interested in understanding if the volume, and thus, the theoretical net energy output grows faster than the mass as we increase the size, say the major diameter, of a theoretical tokamak.

If so, I expect there would be some break-even point where we could build a reactor so large or larger, that the output would be great enough that, if dedicated to thrust, it could accelerate faster than a smaller craft with a smaller reactor.

 

[etudiant]

In space, an ITER could become a propulsion unit if large enough to have plenty of spare flux, if only by removing the neutron absorbing blanket at the point where one would wish to generate thrust.
That should have pretty decent ISP, albeit nowhere near the more exotic designs that have been discussed.

 

[mfb]

I'm interested in understanding if the volume, and thus, the theoretical net energy output grows faster than the mass as we increase the size, say the major diameter, of a theoretical tokamak.

Typically: Yes. That is not a miracle cure to every size problem, however. Another typical behavior is that this approaches a limiting power density.

 

[Xforce]

I'm not sure if you recognize just how little time such an engine would function for a rocket. Unless the ascent profile were altered to try to maximize its use (without tearing apart the rocket), a rocket would spend maybe a minute - the second minute - of a 9 minute ascent in an envelope suitable for scramjet use. They could already be using turbojets for the first minute, in thicker air, if it were viable. I think it is just more trouble than it is worth.

Of course the craft takes off horizontally, a spaceplane. It accelerates to near orbital velocity in the atmosphere, and once it clears the atmosphere, the inertial fusion engines activates and puts it into orbit.
Since fusion engines have very high Isp we expect to reduce fuel and adds more payload