PPPL plasmoid thruster / magnetic reconnection thruster

[edzieba]

PPPL Press release
DOI
Arxiv paper

For this unoptimized high-power case (with a trust [sic] of 50-100N), the ratio of thrust over power is thus about 5-10 mN/kW. We have not yet performed a systematic optimization, but tentatively the optimal parameter range for this new thruster will be ISP (specific impulse) from 2,000 to 50,000 s, power from 0.1 to 10 MW and thrust from 1 to 100 Newtons.

Certainly looks like an interesting concept, flinging out blobs of plasma at a few tens to hundreds of km/s.

[momerathe]

Thanks for the arxiv link! The press release was frustratingly short of useful information.

[momerathe]

Seems similar in concept to the Electrodeless Lorentz Force Thruster. Still, 50,000s Isp is pretty impressive, even if it is only a computer model. I’d  like them to get a NIAC grant or something to take it further.

[punder]

A news article: https://news.sky.com/story/new-concept-for-rocket-thruster-exploits-the-mechanism-behind-solar-flares-12202285

Okay, not much there...

[adrianwyard]

Interesting. A theoretical Isp of 1000-50000 is no joke. Direct link to paper: https://arxiv.org/pdf/2011.04192.pdf

Is there enough there to speculate on how this would be integrated into an actual spacecraft and if there are any 'gotcha's?

+ What are suitable power sources (0.1-10MW), and how heavy are they?
+ What hardware is needed to generate the magnetic fields, and how heavy is it? Superconducting needed?
+ What is the likely fuel, and how heavy is it and the associated tankage?

It sounds like you could have a heavy vehicle, with a (paltry) 100N thrust, but very high Isp. Might be worth running the numbers to see how heavy it could get and retain good transit times to Mars, asteroids, etc...

[philw1776]

At 100 Newton thrust a 100 ton dry mass SpaceX Starship with a crewed 100 ton payload with zero engine mass and zero propellant mass would accelerate very slowly. Less than 1 mm per second. WAY too slow.
There are no 100 MW spaceflight power supplies or even 1 MW ones. Best so far is Kilopower at 1KW, 1,000 times smaller than 1 MW and Kilopower won't fly for years.

[Stan-1967]

Per the Arvix paper, I don't think it is helpful for the scheme to use relatively low voltage ( hundreds of volts) and high current ( >100kA).  I don't see how it really differentiates itself from other plasma thruster schemes when it comes to the problems of producing the power to create a strong enough magnetic field to do work on the plasma.  The power problem for this scheme seems compounded by the pulse duty cycle, meaning this will likely require large & heavy capacitor banks, superconducting materials for the high current/low voltage, & thermal management to radiate waste heat to space. 

Finally, I'm not so clear on what was novel about this idea that make it better that previous similar themes of the idea?  All that is mentioned is novel configuration of creating instabilities in static electrical & magnetic fields that create the plasmoid.  There are two pages of referenced prior work on related subjects that this idea is built on. 

[adrianwyard]

The author mentions that the ability to vary power output is novel, but from my armchair the ability to reduce thrust from a max of 100 N is not that great a feature  But seriously I suppose lower thrust could conceivably optimize power/weight but then you're closing on other simpler and lighter forms of electric propulsion.

[edzieba]

At 100 Newton thrust a 100 ton dry mass SpaceX Starship with a crewed 100 ton payload with zero engine mass and zero propellant mass would accelerate very slowly. Less than 1 mm per second. WAY too slow.

That gains you ~1km/s every ~11 1/2 days of engine operation (278 hours). That's not bad at all when it comes to very high ISP engines.

[philw1776]

At 100 Newton thrust a 100 ton dry mass SpaceX Starship with a crewed 100 ton payload with zero engine mass and zero propellant mass would accelerate very slowly. Less than 1 mm per second. WAY too slow.

That gains you ~1km/s every ~11 1/2 days of engine operation (278 hours). That's not bad at all when it comes to very high ISP engines.

True
But my example used a ridiculous low starting mass addressed at crewed missions to Mars as touted in the articles.

This could be great for un-crewed outer solar system missions given the unicorn megawatt power supply at ? mass.
What's the ISP, thrust, Megawatts power needed vs VASIMR which is at least now a lab device after decades?

[Nilof]

Based on the simulations above, we find thatthere are fundamentally several advantages of this novel thruster, including: 1- High andvariable exhaust velocity as large as 500km/s with injected poloidal field of 500-600G.2- Large and scalable thrust – depending on the size of plasmoid and magnetic fieldstrength, the thrust can range at least from a tenth of a Newton to tens of Newtons. Asthe reconnecting plasmoids leave the device at the Alfven velocity, the thrust scales asmagnetic field squared. 3- The thrust does not ideally depend on ion mass, so plasmacan be created from a wide range of gases, including gases extracted from asteroids.

A big win of this thruster imho that it seems to share with ELF thrusters seem to be the capability to use ISRU gases as propellant, except that unlike ELF thrusters it also has a wide variable Isp range. So you could run it at a higher ISP to get the mining vessel to an asteroid efficiently, and then run it at low ISP using ISRU propellant outgased from the asteroid to bring it to its destination.

The super high end of the Isp range is generally a gimmick for most variable Isp thrusters. I suppose that makes it usable for outer solar system exploration, but there's a ton of competing viable thruster designs at the high Isp range where ionization energy stops mattering to the efficiency and outer solar system thrusters would be optimized for their task anyway.

The main question I would have about them in the lower Isp ranges would be what their footprint is. Hall effect thrusters are flat and easy to mount on thrust plates that fold out. Can this engine do the same or is it bulky like VASIMR? Other than that, the paper seems to assume a thruster radius of roughly a meter, which would still be very compact for a 10 MW thruster and allows for a cluster of them.

At 100 Newton thrust a 100 ton dry mass SpaceX Starship with a crewed 100 ton payload with zero engine mass and zero propellant mass would accelerate very slowly. Less than 1 mm per second. WAY too slow.
There are no 100 MW spaceflight power supplies or even 1 MW ones. Best so far is Kilopower at 1KW, 1,000 times smaller than 1 MW and Kilopower won't fly for years.

Wrong way to think about it. You would put it on a dedicated in-space spacecraft, not bolt a single one onto a chemical stage. Meaning that in the inner solar system you could definitely give it enough power with large solar arrays given that an alpha 1-2 kg/kW is doable with next gen solar arrays, and you would of course size the number of engines after the size of the ship & payload. You wouldn't complain that the Raptor has too little thrust because a single one can't lift the super heavy off the ground by itself.

[rakaydos]

"Third, unlike current thruster concepts that rely on electric fields, the magnetic fields in Ebrahimi’s concept allow the plasma inside the thruster to consist of either heavy or light atoms. This flexibility enables scientists to tailor the amount of thrust for a particular mission. “While other thrusters require heavy gas, made of atoms like xenon, in this concept you can use any type of gas you want,” Ebrahimi said. Scientists might prefer light gas in some cases because the smaller atoms can get moving more quickly."

Interesting. If this can be fueled up from methalox depots, you might be able to have a unified high-impulse-lift/inspace-ISP fuel system, between this and Raptor.

[Nilof]

Re-reading the paper, another nice thing about it seems to be the fact that it can gear all the way down to 1000s Isp. If it stays efficient in that regime and efficiency is less impacted by the ionization energy than Hall effect thrusters, then it could be effective even as an outright replacement for Halls. Depends on how good the performance numbers will be when it reaches higher TRL levels.

Downside seems to be that its power processing unit may end up being more complex than what you would need for a HET, due to the pulsed load

[philw1776]

Based on the simulations above, we find thatthere are fundamentally several advantages of this novel thruster, including: 1- High andvariable exhaust velocity as large as 500km/s with injected poloidal field of 500-600G.2- Large and scalable thrust – depending on the size of plasmoid and magnetic fieldstrength, the thrust can range at least from a tenth of a Newton to tens of Newtons. Asthe reconnecting plasmoids leave the device at the Alfven velocity, the thrust scales asmagnetic field squared. 3- The thrust does not ideally depend on ion mass, so plasmacan be created from a wide range of gases, including gases extracted from asteroids.

A big win of this thruster imho that it seems to share with ELF thrusters seem to be the capability to use ISRU gases as propellant, except that unlike ELF thrusters it also has a wide variable Isp range. So you could run it at a higher ISP to get the mining vessel to an asteroid efficiently, and then run it at low ISP using ISRU propellant outgased from the asteroid to bring it to its destination.

The super high end of the Isp range is generally a gimmick for most variable Isp thrusters. I suppose that makes it usable for outer solar system exploration, but there's a ton of competing viable thruster designs at the high Isp range where ionization energy stops mattering to the efficiency and outer solar system thrusters would be optimized for their task anyway.

The main question I would have about them in the lower Isp ranges would be what their footprint is. Hall effect thrusters are flat and easy to mount on thrust plates that fold out. Can this engine do the same or is it bulky like VASIMR? Other than that, the paper seems to assume a thruster radius of roughly a meter, which would still be very compact for a 10 MW thruster and allows for a cluster of them.

At 100 Newton thrust a 100 ton dry mass SpaceX Starship with a crewed 100 ton payload with zero engine mass and zero propellant mass would accelerate very slowly. Less than 1 mm per second. WAY too slow.
There are no 100 MW spaceflight power supplies or even 1 MW ones. Best so far is Kilopower at 1KW, 1,000 times smaller than 1 MW and Kilopower won't fly for years.

Wrong way to think about it. You would put it on a dedicated in-space spacecraft, not bolt a single one onto a chemical stage. Meaning that in the inner solar system you could definitely give it enough power with large solar arrays given that an alpha 1-2 kg/kW is doable with next gen solar arrays, and you would of course size the number of engines after the size of the ship & payload. You wouldn't complain that the Raptor has too little thrust because a single one can't lift the super heavy off the ground by itself.

My example was chosen to illustrate the point of unsuitability for fast Mars transit as cited previously.
Of course the best use of this conceptual engine is with dedicated smaller mass in-space craft.

[rakaydos]

PPPL Press release
DOI
Arxiv paper

For this unoptimized high-power case (with a trust [sic] of 50-100N), the ratio of thrust over power is thus about 5-10 mN/kW. We have not yet performed a systematic optimization, but tentatively the optimal parameter range for this new thruster will be ISP (specific impulse) from 2,000 to 50,000 s, power from 0.1 to 10 MW and thrust from 1 to 100 Newtons.

Certainly looks like an interesting concept, flinging out blobs of plasma at a few tens to hundreds of km/s.

Plugging in 100 of the largest planned Kilopower reactors (1.5 tonnes each, for 10 kw each)= 150 tonnes of power source for 1 megawatt. Assuming that generates 10 newtons of thrust (and the engine and fuel tank's mass is negligible in comparison) that's an acceleration of 4 millimeters per second every minute of operation. At a mass ratio of only 2 and an exaust velocity of 250 kms, that's 173.3 km/s of Dv.

270 tonnes dry (a starship with 120 tonnes structure) and 1470 tonnes wet (fully fueled starship with standard starship propellants) at the same exaust velocity is 423.6 km/s of delta v.

[jketch]

PPPL Press release
DOI
Arxiv paper

For this unoptimized high-power case (with a trust [sic] of 50-100N), the ratio of thrust over power is thus about 5-10 mN/kW. We have not yet performed a systematic optimization, but tentatively the optimal parameter range for this new thruster will be ISP (specific impulse) from 2,000 to 50,000 s, power from 0.1 to 10 MW and thrust from 1 to 100 Newtons.

Certainly looks like an interesting concept, flinging out blobs of plasma at a few tens to hundreds of km/s.

Plugging in 100 of the largest planned Kilopower reactors (1.5 tonnes each, for 10 kw each)= 150 tonnes of power source for 1 megawatt. Assuming that generates 10 newtons of thrust (and the engine and fuel tank's mass is negligible in comparison) that's an acceleration of 4 millimeters per second every minute of operation. At a mass ratio of only 2 and an exaust velocity of 250 kms, that's 173.3 km/s of Dv.

270 tonnes dry (a starship with 120 tonnes structure) and 1470 tonnes wet (fully fueled starship with standard starship propellants) at the same exaust velocity is 423.6 km/s of delta v.

400km/s is a lot of delta v, but if the acceleration is 4mm/s/min, it will take 100,000,000 minutes (200 years) to use all of it.

[Asteroza]

So, the ELF equivalent of a VASIMR with variable ISP, plus can use ISRU gases. Wonder if nitrogen works...

[edzieba]

PPPL Press release
DOI
Arxiv paper

For this unoptimized high-power case (with a trust [sic] of 50-100N), the ratio of thrust over power is thus about 5-10 mN/kW. We have not yet performed a systematic optimization, but tentatively the optimal parameter range for this new thruster will be ISP (specific impulse) from 2,000 to 50,000 s, power from 0.1 to 10 MW and thrust from 1 to 100 Newtons.

Certainly looks like an interesting concept, flinging out blobs of plasma at a few tens to hundreds of km/s.

Plugging in 100 of the largest planned Kilopower reactors (1.5 tonnes each, for 10 kw each)= 150 tonnes of power source for 1 megawatt. Assuming that generates 10 newtons of thrust (and the engine and fuel tank's mass is negligible in comparison) that's an acceleration of 4 millimeters per second every minute of operation. At a mass ratio of only 2 and an exaust velocity of 250 kms, that's 173.3 km/s of Dv.

270 tonnes dry (a starship with 120 tonnes structure) and 1470 tonnes wet (fully fueled starship with standard starship propellants) at the same exaust velocity is 423.6 km/s of delta v.

400km/s is a lot of delta v, but if the acceleration is 4mm/s/min, it will take 100,000,000 minutes (200 years) to use all of it.

4mm^s^-2 still means you can go from LEO to TMI (lets call it ~5km^s-1 for simplicity) in 1,250,000s, or a little over two weeks. That's not too shabby. But it also means if you want to go to Pluto (~8.5km^s-1), you can burn for 3 and a half weeks on your way out, then another 5 weeks at the end to enter Pluto orbit, and then do the whole thing in reverse to get back home, on 'one tank of gas'. And that's just for a basic Hohmann, you can potentially fly brachistochrone and never shut down at all, just flip over halfway. Being able to sustain a few mm^s-2 for years at a time means you can go damn near anywhere in the solar system in reasonable timeframes and not even need to care about propellant availability at your destination(s).

[rakaydos]

PPPL Press release
DOI
Arxiv paper

For this unoptimized high-power case (with a trust [sic] of 50-100N), the ratio of thrust over power is thus about 5-10 mN/kW. We have not yet performed a systematic optimization, but tentatively the optimal parameter range for this new thruster will be ISP (specific impulse) from 2,000 to 50,000 s, power from 0.1 to 10 MW and thrust from 1 to 100 Newtons.

Certainly looks like an interesting concept, flinging out blobs of plasma at a few tens to hundreds of km/s.

Plugging in 100 of the largest planned Kilopower reactors (1.5 tonnes each, for 10 kw each)= 150 tonnes of power source for 1 megawatt. Assuming that generates 10 newtons of thrust (and the engine and fuel tank's mass is negligible in comparison) that's an acceleration of 4 millimeters per second every minute of operation. At a mass ratio of only 2 and an exaust velocity of 250 kms, that's 173.3 km/s of Dv.

270 tonnes dry (a starship with 120 tonnes structure) and 1470 tonnes wet (fully fueled starship with standard starship propellants) at the same exaust velocity is 423.6 km/s of delta v.

400km/s is a lot of delta v, but if the acceleration is 4mm/s/min, it will take 100,000,000 minutes (200 years) to use all of it.

4mm^s^-2 still means you can go from LEO to TMI (lets call it ~5km^s-1 for simplicity) in 1,250,000s, or a little over two weeks. That's not too shabby. But it also means if you want to go to Pluto (~8.5km^s-1), you can burn for 3 and a half weeks on your way out, then another 5 weeks at the end to enter Pluto orbit, and then do the whole thing in reverse to get back home, on 'one tank of gas'. And that's just for a basic Hohmann, you can potentially fly brachistochrone and never shut down at all, just flip over halfway. Being able to sustain a few mm^s-2 for years at a time means you can go damn near anywhere in the solar system in reasonable timeframes and not even need to care about propellant availability at your destination(s).

Not 4mm/second squared. 60 times worse.

[Nilof]

PPPL Press release
DOI
Arxiv paper

For this unoptimized high-power case (with a trust [sic] of 50-100N), the ratio of thrust over power is thus about 5-10 mN/kW. We have not yet performed a systematic optimization, but tentatively the optimal parameter range for this new thruster will be ISP (specific impulse) from 2,000 to 50,000 s, power from 0.1 to 10 MW and thrust from 1 to 100 Newtons.

Certainly looks like an interesting concept, flinging out blobs of plasma at a few tens to hundreds of km/s.

Plugging in 100 of the largest planned Kilopower reactors (1.5 tonnes each, for 10 kw each)= 150 tonnes of power source for 1 megawatt. Assuming that generates 10 newtons of thrust (and the engine and fuel tank's mass is negligible in comparison) that's an acceleration of 4 millimeters per second every minute of operation. At a mass ratio of only 2 and an exaust velocity of 250 kms, that's 173.3 km/s of Dv.

270 tonnes dry (a starship with 120 tonnes structure) and 1470 tonnes wet (fully fueled starship with standard starship propellants) at the same exaust velocity is 423.6 km/s of delta v.

Kilopower is NOT intended as a low-mass electricity source for electric propulsion. It's an RTG replacement. Its barely even competitive with advanced solar at Jupiter's distance from the sun.

What you should be looking at is roll-out solar arrays in the near term, and various setups for thin-film solar arrays on tensegrity structures in the long term. Those enable masses of less than 1 kg/kW at 1 AU in the medium to near term, even scaled up to hundreds of megawatts.

For a rough example of what can be done at those power levels, this entry in the rocketpunk manifesto is a classic:
http://www.rocketpunk-manifesto.com/2008/05/on-gossamer-wings.html

If you have variable specific impulse you can do a bit better, see the formula in my signature that replaces the rocket equation for constant-thrust trajectories.