"The essential idea is that air is ionized to a plasma, which is rapidly heated and allowed to expand to generate thrust."
So this is just a conventional heat engine, with an electric heater. This heater may be able to get the air much hotter than other methods, but the thing about a heat engine is that you cannot get useful work out unless the working fluid can expand sufficiently. An afterburner creates more thust by heating the air to a higher temperature than could be tolerated by the turbine, but the air is by then at a relatively low pressure, and so the Carnot efficiency is very poor - most of the extra fuel's energy goes into producing a hotter (and very visible) exhaust plume.
So, for this to be a component of a jet engine, it will need a compressor comparable to, or with an even higher presure ratio, than in current jet engines, and that compressor will have to be powered somehow (IIRC to the tune of about 50,000 SHP in the biggest engines now in use.)
For the most part, it makes no sense to use electricity to power a heat engine. In guessing where this might be useful, the only scenario I have come up with is for hypersonic ramjets, where electric motors turning fans are not an alternative, and possibly especially on worlds where the atmosphere does not support combustion.
Yeah, it seems like it might make more sense if they could figure out a way to use the RF fields to actually accelerate some of the ions in the plasma. Then you might be able to reach much higher exhaust velocities than with a conventional jet engine. I don't think your thermodynamic objections would apply to that case since the energy would remain organized, it wouldn't just be heat.
To do that though you'd need to find a way to maintain overall electrical neutrality.
>Yeah, it seems like it might make more sense if they could figure out a way to use the RF fields to actually accelerate some of the ions in the plasma.
That's essentially what they are doing, but in an oscillating path rather than a continuous path:
In the waveguide, the charged particles in the plasma start to oscillate with the microwave field (aka: RF) while rapidly heating. The ions, atoms, and electrons collide with each other frequently, spreading the energy from the ions and electrons to the neutral atoms, heating the plasma rapidly. As a result, the researchers claim that the plasma rapidly heats to well over 1,000°C.
I don't know if you can rectify RF energy like you can with voltage. That might be one way to create a more continous path. They might also be able to add an electromagnetic field to create a net velocity out the provebial barn door as they do with ion thrusters, but I think the problem there is you just can't get enough free air path to accelerate to very high speeds...it'd be like trying to drive an F1 car at full throttle in bumper to bumper traffic. You could also pass a current through the plasma to generate lorentz forces like they do in plasma driver rail guns.
One huge advantage with this design is no moving parts and presumably extremely low manufacturing costs. So it might not be awesome for commercial airliners but it could be useful for high endurance UAVs. Pack a few dozen of them and use just the ones you need.
> I don't know if you can rectify RF energy like you can with voltage.
You can in essence, because charged particles have inertia, they don't instantly follow the EM field. That's how RF particle accelerators work. But there you're working with a beam of charged particles, not a neutral plasma.
It still sounds to me that they are essentially using the RF energy just to heat the plasma and in that case the parent's objections seem valid, though I don't know much about jet engine design.
A few months ago, there was a demonstration of a lightweight model airplane propelled by the air currents produced by a corona discharge [1], and ion rocket motors have great specific impulse, but they are both low-thrust devices. I believe you are right, that these are not heat engines, any more than an electric motor is: in all these cases, the moving part is directly accelerated in the direction we want it to go.
> For the most part, it makes no sense to use electricity to power a heat engine.
Exactly. And modern aircraft engines generally just use the turbine to power a high bypass fan, at which point why pretend to be a heat engine when you could just spin the fan directly and save a tremendous amount of power?
Modern high bypass turbines provide a lot of power for the weight and extreme efficiency. Essentially, they are a gas turbine which then gets to extract extra energy from their exhaust gasses without adding a lot of weight or mechanical complexity.
There are a lot of trade offs involved, but for large 500+ MPH aircraft high bypass turbofans are simply the most cost effective option.
Yeah, exactly. Being able to stick the turbine in the middle of the ducted fan and then use its exhaust for extra thrust while directly driving the fan with the generated gases is the cherry on top.
If we were going to replace a high bypass turbofan with anything, though, I'd expect it to be a high power AC motor directly driving a ducted fan. Realistically for long haul air travel, though, my guess is we'll stick with turbofans and just use biodiesel or something.
An idea I've had is to put a small generator instead of a shaft and gearbox, and put the AC motor powered fan somewhere else. Could improve aerodynamics, stealth properties etc.
Is your point that they must necessarily be heavy? Because I don't think so. Modern electric motors are very compact and powerful.
Also you'd not need gearbox and shafts. You can completely decouple the fan from the turbine. You could put the fan downwards and the turbine back for instance, or the turbine on the top to hide its heat signature somewhat.
Doing some back of the envelope calculations to see if we're in the right ball park here:
A Tesla motor weighs in at 70lbs (31kg) on its own, not including the inverter or any gearbox, and generates 362hp. Or 11hp / kg.
Using a turboprop because it's easier to do a straight hp comparison, the PW150 has a dry weight of 716.9kg and produces 5000hp continuous. Or about 7hp / kg.
Of course it's absolutely not a fair comparison for a number of reasons, for a start running a Tesla motor at full power for more than a few minutes at a time is going to result in an overheated motor where as the PT150 can do that all day long. The Tesla motor weight doesn't include the inverter which would be needed (unless it's all designed to run each motor at a synchronous speed with the generator...), nor the cooling system, a gearbox (the low pressure turbine of a high bypass engine is only ~4000rpm, the Tesla motor does 18000rpm). Not to mention the elephant in the room, you still have the weight of the generator plus the prime mover (or battery...).
To be honest that surprised me. I think the Tesla motor is slightly misleading. If you've ever seen a 3hp industrial motor you'll see where I'm coming from, they are about the same size as the Tesla's motor and probably weigh twice as much. I appreciate that the tech in those motors is very old compared to the Tesla, but if they are that far out you'd think there would be profit somewhere in improving them.
The difference here (as often is the case when comparing industrial vs. consumer equipment) is that the Tesla motor is rated for peak load under favourable conditions while the industrial motor is rated for continuous load under worst-case conditions. That 3hp industrial motor will run at 3hp shaft power at its maximum rated temperature for its rated operational life.
Many DIY electric car conversions use DC motors rated between 9hp and 20hp. They routinely get 100-200hp+ out of these motors for a few seconds at a time.
The industrial motors don't have permanent magnets, so are cheaper to build and do not run any risk of permanent magnets become de-magnetized. Downside is they are very heavy.
Replacing chemical combustion with electrical heating in a heat engine was an idea that popped into my head a number of years ago while thinking about the heat exchanger for Skylon's SABRE engine. I'll skip the drunken, derailed train of thought that made that particular leap, but off and on since then I've been really intrigued by it, and spent more time thinking about it than I'd care to admit!
First and foremost, I agree that the compressor is probably the most difficult part (though theoretically, if you pre-ionized the gas, you could use magnetic compression). Also, I completely agree that, given current energy densities for electric storage, this is only something that could be useful in some really niche applications.
But that being said, some of those applications are really cool! For example, one of the major challenges of VTOL aircraft is that the rotational inertia of turbines is so great that it's very, very difficult to rotate them during transition from vertical to horizontal flight. Something like this would massively decrease the rotational inertia, making it much simpler mechanically to create tiltwing aircraft.
Also, my understanding is that typically, conventional jet engines are limited primarily by the maximum temperature limit of the turbine blades. Because your compressors here would have to be powered by electricity as well (nothing else makes any sense!), there's absolutely no reason to have a turbine at all; you'd just want a plain old expansion nozzle. That means you could pump way more heat into your plasma, making your engine much more power dense. In other words, your engine could be potentially much smaller for the same thrust, which would be a big deal. Turbine blades need to be both very strong due to their rotational velocity, and extremely temperature resistant because they're literally sitting in the exhaust of a jet engine, which makes them not only really expensive, but also very, very challenging from a metallurgical perspective.
Another, potentially very interesting, application is if you have too little oxygen in your atmosphere to support combustion -- for example, on Mars. Sure, we're about to send a mini electric rotorcraft there, but the atmosphere makes it really very challenging to do that, because the classic "my rotor tips are too close to the speed of sound" problem is much, much more difficult there.
Any kind of ramjet, as you mention, is a possibility, but this would also make it a lot easier to make transition engines (like the J58 that powered the Blackbird) that start as a conventional compressor-fed jet engine and, at cruise speed, transition into a ramjet.
Regardless of application, this is such a fundamental change to the design limitations of jet engines that a lot of the usual design logic simply doesn't apply anymore. Thermodynamics are infamously complicated, which makes it really difficult to draw performance comparisons between an 80-year-old mature technology and something so radically new and different. One way this gets substantially more complicated is that in a traditional jet engine you need to be worried about combustion efficiency, flame stability, etc etc, plus you have to siphon out enough energy to run the engine's compressor, and power the rest of the aircraft (likely indirectly, through an APU!). All of those take a big efficiency hit in traditional engines, whereas this would be, nominally, much better. So my gut would be that, all other things being equal, the powerplant on an airplane with an electric jet engine would be both smaller/lighter and more efficient. But again, this is hard to reason about!
I would be ecstatic to see one of these flying around, but don't expect it to end up in a passenger aircraft any time soon or anything. For that, we need better batteries!
> There's absolutely no reason to have a turbine at all; you'd just want a plain old expansion nozzle. That means you could pump way more heat into your plasma, making your engine much more power dense.
I agree that replacing the turbine with an electric motor seems to be the only way to go with this, but the point I am trying to make here is that if you merely increase the temperature of the working fluid without changing the pressure ratio, you will get some increase in thrust, but at the cost of a worse Carnot-cycle efficiency: quite a bit of the additional energy input goes to waste in the form of a hotter exhaust, because it cannot be expanded enough to convert it to useful work.
So can we increase the pressure ratio? if it were feasible to do so with current technology, we would already be doing so, as combustion jet engines would also benefit from increasing it. When comparing plasma and combustion jet engines, we must assume that both will be operating at the highest feasible pressure ratio.
That does not automatically rule out this technology, as it may offer something in trade-off for its limited efficiency, but in the case of electric propulsion, the storage options are currently so limited that efficiency is highly valued.
The question to be answered is this: for a given electricity source, will this give me anything of value over using all the power in an electric motor driving a fan? For subsonic flight, I am very skeptical that it can even come close to having anything to offer.
The J58 is often described as a hybrid turbo-ramjet, but that is hype to some extent: it is a low-bypass turbojet with an afterburner and a pressure-recovery intake, but that describes every supersonic airplane. It is the pressure-recovery inlet that makes all these engines somewhat ramjet-like, and in the J58 there is just more of it. The distinction is a matter of degree; even subsonic jets take advantage of pressure recovery.
Pressure recovery does two things: it increases the overall pressure ratio, and it slows down the inflow to the compressor to below supersonic speed. The former is only a benefit for heat engines (Carnot efficiency, again.) Therefore, I think the only reason for having such an inlet in an electric-fan jet engine is if a supersonic fan is infeasible, and they may well be. If so, then it may be the case that makes sense to use some of the available electric power to heat the compressed flow downstream of the fan, but it is not obvious to me that this would be a better use of that power than using it all in a bigger fan.
I see from here [1] that supersonic compressors, and therefore presumably fans, are feasible, though have not been very successful (maybe because pressure recovery is a better option for heat engines.)
By using electric power in a heat engine rather than in a non-thermal process, you are already committed to throwing about two-thirds of it away, so there have to be some quite compelling benefits elsewhere to make it a net win overall.
While it is certainly true that any device that heats slowly would be impractical on that basis alone, I do not think rapid heating provides any relief from the laws of thermodynamic efficiency. Combustion heats very rapidly, though if this device is even faster, perhaps that would be significant in hypersonic applications? I believe one of the problems of supersonic-combustion ramjets is maintaining a stable flame.
I keep coming back to hypersonic ramjets, not because I think they are the perfect application, but because I am not sure there is any application in an otherwise conventional jet engine - i.e. one involving a rotary fan or compressor - where you would not be better off using all the available electric power in an electric motor to spin a somewhat larger fan, and not bother with heating at all.
Update: Maybe, for replacing the turbojet engines of supersonic airplanes, you need heat to reaccelerate the flow above supersonic speeds, in a convergent-divergent nozzle? But supersonic wind tunnels seem to achieve those speeds without heating the flow? I have wandered well into my area of ignorance here...
If the article is right about the scaling required then I'm very skeptical that this will ever be practical. A 1 kW magnetron is a part found in just about every microwave oven. Scaling that up by 4 orders of magnitude would need a 10 MW RF source or amplifier. I'm pretty sure those don't exist and if they did they'd be very large and heavy. A long time ago I used to do RF engineering for particle accelerators and the most powerful continuous wave RF amplifiers I ever heard about were on the order of 1 MW.
An even bigger problem would be efficiency, which this article doesn't even mention (haven't looked at the original paper). High power RF amplifiers aren't particularly efficient, I would guess around 30%, and there would also be waveguide losses and cavity losses if any resonant effects are used to get high enough electric fields. I would be surprised if there's much hope of that competing with conventional jet engines on efficiency.
> I'm pretty sure those don't exist and if they did they'd be very large and heavy
I'm clueless on these things, but what order of magnitude for "heavy" are we talking about here? If we take a 1Kg microwave and add 4 orders of magnitude we're at 10,000Kg and a 747 weighs in at ~200,000Kg, so by that measures it seems achievable. I'd also assume existing terrestrial ones aren't optimized for weight in any way.
> I would be surprised if there's much hope of that competing with conventional jet engines on efficiency.
Efficiency isn't the only measure, there if energy can be cheaper than fuel then less efficient can win out. There may also be applications for long running, low weight flight powered by solar like starlink.
> Scaling that up by 4 orders of magnitude would need a 10 MW RF source or amplifier. I'm pretty sure those don't exist and if they did they'd be very large and heavy.
Probably a dumb question, but it's it just a matter of more power and cooling for the magnetron? I'm thinking the size of the magnetron is determined by the wavelengths you're trying to produce. (I don't mean to diminish the challenge of applying this tech, a 10 MW power source would still be quite large for an airplane.)
Yes, I think so. If the frequency is high enough so that it's not the limiting factor the size would mainly depend on how much power you can dissipate in a given volume.
The highest power RF amplifiers I've ever seen were the klystrons used for the LEP and ESRF accelerating cavities. Those were large in part because of the relatively low frequency of 350 MHz but I believe there was also a large water-cooled absorber that dissipated whatever power was left in the MW+ electron beam given the conversion efficiency.
I'm sure it will be a long road to practicality, but that it ionizes all of the air and drives it out is fascinating. Of course it also then requires strong electric drivers since it's basically driving the engine with the magnetron.
I can kind of imagine improved transistors and other technology making scaling the magnetron easier, hopefully all that work on fusion containment has helped us understand plasma better here too!
That comparison wouldn't be as bad you might think.
The first functional Diesel engine for example had about 13kW of power and was only about 5x larger than a modern car engine.
This was in 1896. From 1923 onward, the engine type was small and powerful enough to be used in lorries (trucks).
The first petrol engines compare even better to modern engines (in terms of power-to-weight ratio).
When you say "continuous", would you consider some very high frequency solid state switching amplifier to be "continuous" enough for this application? I realize it's a different order of magnitude, but those GaN/Si transformers make me wonder if we aren't far off from some kind of megawatt scale solid state amplifier shakeup...
It's conceivable. I've seen solid state RF amplifier modules successfully combined into some quite high power units and that was over 20 years ago. I haven't followed the technology since so I'm not up to date on that at all. In any case I think a 10 MW amplifier and associated power supplies and cooling equipment would need to be large and heavy.
A single jet engine generates about 20 MW of power and I'd be very surprised if you could pack 10 MW of RF amplifier into that kind of size and weight envelope.
Makes sense. Maybe we'll see a fusion plasma thruster before the power/energy storage shrunk enough to make this viable as is (in the atmosphere and gravity of Earth at least).
Most high power microwave devices aren't build on semiconductors like Silicon (solid state), yet. They are normally valves of some sort, like magnetrons, klystrons or travelling wave tubes.
Some radars would be in the mega-watt range, but only pulsed with low duty cycles or 1%.
It sounds like the technology hasn't evolved much in the last 20-30 years then.
Around 1995 I was involved with a project where an engineer designed and built a multi-kW CW (I don't recall the exact value) pre-amp out of transistor based modules. There was a plan to apply the same approach to a later project at much higher power levels (100's of kW's) where the conventional approach would have been to use a klystron. I left that team (and the field) though so I don't know if the larger version was ever successfully implemented.
EDIT: I should note that this wasn't in the microwave regime, it was 500 MHz.
Submitted title was "Wuhan scientists develop jet propulsion by microwave air plasma", which broke the site guidelines by adding linkbait. Submitters: please don't do that. https://news.ycombinator.com/newsguidelines.html
How is this better than an electric motor turning a fan?
Modern high-bypass jet engines get the vast majority of their thrust from turning a fan. The turbine part is (mostly) just used to generate torque to drive the fan.
Modern electric motors also have ridiculously high efficiencies (> 97% isn't uncommon).
So how would using electricity to heat the air be better than using the same electricity to turn a fan? The only place I can think of is high supersonic where fan efficiency starts to drop.
This makes me think of all those UFO sightings of weird craft that flash multicolored lights that look like emission spectra you would get from plasma. I wonder if at least a few of these might have been classified experimental propulsion systems similar to this. The basic physics of this is not new and it's not like these programs have lacked the funding to experiment with crazy tech.
Correct me if I'm wrong but would this not spew massive amounts of nitrogen oxides? You know, the nasty pollutants you get when superheating air, such as in fuel-efficient diesel engines (see Dieselgate).
I think those would burn, since these engines are meant to breath air (with oxygen in it.) Tungsten filaments work in lightbulbs because those are filled with an inert gas, or vacuum. The filaments burn quick when the bulb is broken.
I don't think that's necessarily a useful question at this point, because some of the basic thermodynamic assumptions don't work out for deciding between batteries vs. using a fuel-burning generator. Normally, it's pretty obvious that "Hey, if we're going to burn hydrocarbons to produce thrust, we might as well do it directly and not deal with necessary conversion efficiency losses between heat -> electricity -> heat." Since they're dealing with RF-generated gas plasmas, though, there might be efficiency gains with acceptable configurations and lower intermediate losses, etc. over traditional systems that makes those conversions acceptable. I've worked in a lab accelerating gas plasmas using helicon antenna, and as the article mentions, scaling those systems isn't a trivial implementation detail. It's very tough to reason about them from first principles, or even small-scale models because of the non-linearity in the relevant physics.
So, I guess in summary, imho it's probably a bit too early to talking about what kind of battery would be used, since it's unclear from the current system that a usable version would actually use batteries vs. another form of energy storage.
How is this diffferent from the antigravity patent?
"a device that uses a microwave emitter to create a high-frequency electromagnetic wave through a cavity to create a polarized vacuum. This polarized vacuum, in turn, reduces the mass of the vehicle containing the device."
There is a massive working medium. For massless particles such as photons the connection between energy E and momentum p is E = p *c or p = E/c. Since c is large, the momentum p you can carry per energy E is small. For a massive particle with mass m the connection is given by E = p^2 / (2 m) or p = sqrt(2 m E). As long as E is small (so small that the velocity of the particle is small compared to the speed of light, but that is required anyway for the non-relativistic approximation I am using here a) the sqrt doesn't hurt you too much and you can actually carry much more momentum per particle.
In essence, this is a jet with a unique compressor, using microwaves to create high temperatures rather than burning fuel. The benefit is that the reaction mass can be entirely normal atmospheric air.
So this is just a conventional heat engine, with an electric heater. This heater may be able to get the air much hotter than other methods, but the thing about a heat engine is that you cannot get useful work out unless the working fluid can expand sufficiently. An afterburner creates more thust by heating the air to a higher temperature than could be tolerated by the turbine, but the air is by then at a relatively low pressure, and so the Carnot efficiency is very poor - most of the extra fuel's energy goes into producing a hotter (and very visible) exhaust plume.
So, for this to be a component of a jet engine, it will need a compressor comparable to, or with an even higher presure ratio, than in current jet engines, and that compressor will have to be powered somehow (IIRC to the tune of about 50,000 SHP in the biggest engines now in use.)
For the most part, it makes no sense to use electricity to power a heat engine. In guessing where this might be useful, the only scenario I have come up with is for hypersonic ramjets, where electric motors turning fans are not an alternative, and possibly especially on worlds where the atmosphere does not support combustion.