Megawatt Class Direct Fusion Drive For 25X ISP and 3 Times Faster to Pluto

Princeton Satellite has had various NASA, SBIR and IR&D grants to develop a multi- megawatt-class nuclear fusion propulsion and space power system. This funding has enabled them to precisely simulate their designs and performed experiments. They have stated they would need $100 million and five years to actually make a full megawatt propulsion system. They have not received the level of funding needed to proceed with the main development.

Studies of electron heating in PFRC-2 surpassed theoretical predictions and recently reaching 500 eV with pulse lengths of 300 ms, and experiments to measure ion heating with input power up to 200 kW are ongoing. When scaled up to achieve fusion parameters, PFRC would result in a 4-8 m long, 1.5 m diameter reactor producing 1 to 10 MW.

They have published various journal papers that review their computer models and simulations of the system. The Direct Fusion Drive concept is an extension of ongoing fusion research at Princeton Plasma Physics Laboratory dating to 2002.

The NASA NIAC project included analysis of the following subsystems: the superconducting coils, heat extraction system, startup system, radiators, and shielding.

Direct Fusion Drive (DFD) would produce between 5-10 Newtons thrust per each MW of generated fusion power, with a specific impulse (Isp) of about 10,000 seconds (chemical is 300-400ISP) and 200 kW available as electrical power. These would be first-generation capabilities.

The DFD system would have
* 35% of the fusion power for thrust
* 30% to electric power,
* 25% lost to heat, and
* 10% is recirculated for the RF heating.

Modeling shows that this technology can potentially propel a spacecraft with a mass of about 1,000 kg (2,200 lb) to Pluto in 4 years. The Pluto Express flyby mission took 9.5 years to reach Pluto. The DFD system would be slowing down and going into orbit around Pluto.

The modeled system would have 2 MW of power for use at Pluto. It could transfer up to 50 kW of power from the orbiter to the lander through a laser beam operating at 1080 nm wavelength.

Better Superconductors Would Make This System More Feasible and Have Higher Performance

Superconducting coils are a major portion of the engine mass. In order to estimate this mass, they reviewed published data on both low-temperature and high-temperature superconductors. Current generation Amperium 12 mm high-temperature wire has a current capacity of 350 A at 77 K, but 700 A at 30 K. This wire has a linear density of 0.2 g/cm. They counted the number of turns to produce 3 MA (3e6 A) in a 0.5 m radius coil: 8572 turns at 77 K and 4286 turns at 30 K, or 579 kg and 298.5 kg, respectively. Considering a single-engine will require 6 to 8 such coils plus the nozzle shaping coils, producing a 1 MW engine on the order of 1000 kg is clearly driven by superconductor mass. High-temperature superconductor companies are working to make thinner tapes with less cladding, but also consider low-temperature superconductors. They need to be cooled to 4.2 K, limiting the choice of coolant and increasing cryostat mass, however, a 1.04 mm NbTi wire has a linear density of just 0.063 g/cm and a capacity of 700 A. The same number of turns of this wire would has a mass of only 84.6 kg. This huge range in available properties is one reason they are still working on superconductor designs.

Nextbigfuture has covered the Princeton Satellite Direct Fusion Drive system several times. The June 2019 article was the last coverage.

Nuclear and Future Flight Propulsion – Modeling the Thrust of the Direct Fusion Drive

Direct Fusion Drive for Interstellar Exploration is paper than covers various possible space missions.

DFD uses an innovative radiofrequency (RF) plasma heating system. The thrust augmentation method is described along with results of multi-fluid simulations that give an envelope of expected thrust and specific impulse. The power balance is described and the subsystems needed to support the fusion core are reviewed. The paper gives the latest results for the system design of the engine, including just-completed work done under a NASA NIAC study. A mass budget is presented for the subsystems.

The paper presents potential interstellar missions.

One is the proposed 550-AU mission that would use the Sun as a gravitational lens for exoplanet research. This mission can be done without a deceleration phase.

Future flyby missions that would need major technological advances and a mature version of the technology would flyby the nearest star.

They sketch a mission to orbit a planet in either the Alpha Centauri A or Alpha Centauri B systems. The mission analyses include a communications system link budget. DFD can operate in an electric-power-only mode, allowing a large fraction of the fusion power to be used for the payload and communications, enhancing the scientific return. All of the missions start in low Earth orbit.

2014 Paper on Space Rapid Transit

Space Rapid Transit (SRT), is a horizontal-takeoff launch vehicle that would revolutionize both the space launch and flight transportation industries. To break the cycle of escalating space launch systems cost, it is necessary to consider concepts that are drastically different from current launch options. SRT is a fully reusable two stage to orbit vehicle. The Ferry Stage is powered by a dual fuel coaxial turbofan ramjet. The turbofan stage uses jet fuel while the ramjet uses hydrogen. The Orbiter uses liquid hydrogen/liquid oxygen engines. Stage separation is at Mach 6.5 at 40 km. The full system, Ferry with reusable Orbiter, is expected to deliver payloads to low earth orbit for less than $300 USD/kg.

NOTE: The SpaceX Falcon Heavy has pricing of about $2200 per kilogram and with a lot of flights and majority reused could get to $1000 per kilogram. The upcoming fully reusable SpaceX Super Heavy Starship could reach $100-1000 per kilogram pricing.

52 thoughts on “Megawatt Class Direct Fusion Drive For 25X ISP and 3 Times Faster to Pluto”

  1. If I remember right, it was a comm laser they used to contact Earth.

    A VERY powerful comm laser.

    But the Kizinti (?) telepaths on board the warship were CERTAIN the ship was unarmed. Nobody had any knowledge of weapons.

    And the laser? Oh, that was just a big flashlight. Heh.

  2. .1. I didn’t get into the EK wars much myself
    .2. But I think Niven didn’t actually write the EK wars. He had a personal rule about never writing war books. Something about he’d never fought in a war, write what you know, stay in your lane etc. He doesn’t let this stop him writing about escaping from alien vampire armies on Megastructures about Alien stars though. (Unless his life is more interesting than wikipedia lets on.)

  3. True. But it was harsh to say Kiwis can’t fly.
    Well… technically Kiwis can’t fly. But they can in free fall.

  4. Yam. I kind of remember that, but didn’t particularly enjoy the EK wars multology. Then again, I din’t much care for Niven’s writing style in general. Oh well.

  5. I think that that is what I said as well. Got to get far enough away so that the close-to-Sol gasses and glowing bits don’t interfere with what ultimately is a quite dim target system many light years yon. 

    The real ‘big problem’ is the nature of the sun-blocking shield. One thing kind of works in our favor: The size of the sol-block shield is precisely defined by how far it is from the imaging camera … the size of Sol … and the distance from Sol. However, often forgotten is that unlike a spherical section lens (i.e. any Newtonian or Schmidt), Sol’s gravitational pull isn’t at all ‘nice to image from’, unless the pass-region is limited to a modestly narrow annulus (ring gap), of only a few hundredths of a solar radii.  

    Then, using the concept of a Fresnel wave-plate, the sun-block can both effectively block Sol, and make up for the aspherical aberration the annulus will be passing from Sol’s gravitational field. 

    I do believe — but I don’t have the math chops to prove — that one simply needs to let the Sol Blocker be further-and-further away from the imaging spacecraft in order to maintain proper focus.  

    However, if focal-distance and Fresnel wave-plate pattern are critically linked, then … well, it becomes kind of a one-sided limitation.

    ⋅-=≡ GoatGuy ✓ ≡=-⋅

  6. Took me a few seconds to figure out how to interpret “J/N*s”. The literal interpretation is “the energy needed to apply 1 Newton of thrust for 1 second” (at the given Isp). But then I realized that it’s also W/N. Which is the power to thrust ratio. Just pointing it out for other readers.

    edit: Also re your PS: E/mv is energy-per-impulse, not impulse-per-energy. That is, how much energy is needed to deliver a unit of impulse.

  7. I think it was Larry Niven who had an SF novel where that was how the plot climax was solved. The humans met the alien attackers who were confident that the humans were unarmed but the sneaky Earthlings just pointed their space ship exhaust at them and went to full power.

    This was of course merely the opening round in the Earth-Kzin wars.

  8. It may well be easier to have a self contained fusion unit with no exhaust.
    But at this point a fusion power plant that did exhaust high velocity ions would still be a major breakthrough that would attract $billions in eager investment.
    But nobody has managed it.

  9. To be fair, New Zealand has launched a number of orbital, if not interplanetary, space missions.
    The private launch company Rocket Lab has launched 48 satellites.

  10. These short fiction stories, written in the style of a news article, where someone speculates about a possible, future, fusion engine; and then spends most of their effort designing spaceships and missions around said science fiction engine, bring out the goat in all of us.

  11. I couldn’t resist. I went biking with the guys and we finished up with a huge meal of bacon and eggs and it was all I could think about.

  12. Nice flame-bait, sad-sack. Try as I might, I cannot recall a single interplanetary mission by … mmm … Nigeria, Yemen, Senegal, Paraguay, Nicaragua, Denmark, Latvia, Azerbaijan, Kazakhstan, Nepal, Indonesia, New Zealand (to spread the wealth), Togo, Vanuatu or Jamaica. As it were. Let us know when that changes a bit. ⋅-=≡ GoatGuy ✓ ≡=-⋅

  13. Mmmm… as others have opined, that just ain’t so. 

    V exhaust, effective = Ve = 9.81 × ISP

    E exhaust = Ee = ½m Ve²
    Ee = ½ (1 kg) • (9.81 ISP)²
    Ee = 0.5 × 96.23 × ISP²
    Ee = 48.12 ISP²

    E/N-s = Ee / ( Ve × 1 kg)
    E/N-s = 48.12 ISP² / ( 9.81 × ISP)
    E/N-s = 4.91 ISP

    So, if ISP is 5,000

    E/N-s = 4.91 × 5,000
    E/N-s = 24,525 joules per newton-second of thrust. 

    This would be the ABSOLUTE MINIMUM energy. Real thrusters take more energy ‘cuz some is lost to visible, IR and UV radiation; more is lost to heat; more still is lost to erosion, ion-ion impacts, and beam heating. Even more, to all the indelicacies of converting low-voltage to high, corona discharge, you name it. 

    So, it takes substantially more than 25 kW to create 1 continuous newton of thrust at 5000 ISP.

    ⋅-⋅-⋅ Just saying, ⋅-⋅-⋅
    ⋅-=≡ GoatGuy ✓ ≡=-⋅

    PS: E/N-s is also E/mv or impulse-per-unit-energy.

  14. The Sun is a spherical lens of infinite size (as far as it’s gravity extends). 550 AU is where photons that just graze the surface come to a focus. Photons that pass 1.5 or 2 radii from the Sun’s center are bent less because gravity is weaker. Therefore they come to a focus farther away. So if you don’t stop, you can use successive focal points from progressively farther misses by photons.

    It’s not practical to use the 550 AU focus, because you get interference from spicules, prominances, and the corona. It’s also hard to exactly block the main disk of the Sun. Most serious proposals assume you will use 800-1000 AU.

    Things are so far apart out there, that you would likely send a dozen or more telescopes in different directions to look at different parts of the sky. We now know of a number of “scattered disk” objects whose orbits reach this region, and there are likely many more our telescopes can’t find yet. They will likely have lots of ices. So our observatories can refuel to keep moving around and and look at other targets.

  15. When it comes to fusion, I am sick of the hype. Build a prototype engine that work and then tell me about it. I can generate hot gas all by my self so that ain’t news.

  16. Lol, do me a favour. Don’t try. Americans can’t do anything except starting a project with a world changing/saving banner.

  17. I just went back and looked, seems like some of my figures were reversed. Instead of 7.7kW/N, it was 24kW/N with Isp of 4190s for the NEXT Ion Thruster. Input power was 7.7kW.

    My VASIMR figures came from another article that claimed that it could do 12,000 Isp. I didn’t think to find out what it’s power-thrust would be at that Isp.

    Lastly, I was groggy when I posted that comment, so I didn’t do much research before posting it…

  18. There is work being done on using meta materials for heat dissipation. If it pans out, it would allow for far more efficient radiators to be designed and built.

  19. They design is much simpler and much, much smaller than giant Tokamaks like ITER, even smaller than the new, compact Tokamak designs by say Tokamak Energy. So 100 million is not THAT unrealistic.

  20. Arguably, a fusion rocket is easier than a fusion reactor (provided the rocket doesn’t need to produce net power, although that doesn’t seem to be the case here). And ITER is highly inefficient in its budget. Still, $100M seems low indeed.

  21. P = F * g0 * Isp (maybe * 0.5, not sure)

    So it’s physically impossible to get more than 10 N/MW at 10000 s Isp (maybe 20 N/MW if there’s that *0.5 factor). You either get less thrust or less Isp. If you want 40 N/MW, you’ll get a lower Isp of no more than about 2500 s (actually less, due to energy losses).

    Similarly, the 24 kW/N figure for VASIMR is at a lower Isp of ~2400 s. At 12000 s Isp, it would need at least 120 kW/N. Probably closer to 200 kW/N, due to losses. In fact, their VX-200 prototype reached an Isp of only 5000 s at 37 kW/N.

    Your claim of 5000 s Isp at 7 kW/N is also bogus. Physics doesn’t work that way. 70 kW/N is more like it.

  22. Let me get this straight.. It’s really, really, difficult to make a commercial fusion reactor, but if you include some plasma thrust in it, it is so easy that all you need is a bit of simulation SW from Princeton..? And of course, 100 million USD is sufficient and not the billions of billions of dollars required for ITER…?

  23. It doesn’t stay focused, and the energy of the individual particles isn’t bad compared to cosmic rays.

    But if you were to launch from LEO, you might fry a few satellites, sure.

  24. Instead of building these fusion drives, let’s just get space-rated fission power generators built.

    There are already space rated Ion Engines with Isp of 5,000s and require only 7kW of power per Newton of thrust.

    VASIMR could reach up to 12,000 Isp, but requires 24kW per Newton. Now, if we could create, say, a 1MW fission reactor, thats 41 Newtons of thrust, *way* more than you’d get from this fusion drive(5N/MW), and similar Isp.

  25. But if you use unicorn bacon and fairy eggs you still get 11.568 kg of bacon and eggs, but you need only 1/10 of the ingredients! I have simulations that confirm it!

  26. Doesn’t the “exhaust” become something like a high-energy particle beam? What happens in the sci-fi future when everyone is zipping around catching a blast from all these focused flows? Just curious.

  27. Hmmm… yes, “a fair bit further” is accurate.

    From my calculations, up to about 5× the distance of criticality.  The 550 AU critical point though is doubly-charmed. IF (a big IF) one has a very, very accurate Fresnel patterned star-shield AND annular ring to block Sol, then the 550 AU would allow a thin sliver surrounding Sol to have the full collecting the power of a telescope no less than 10,000 km² in area.  

    However, with the sliver being that close to the solar limb, it is also the case the all the ‘noise’ of the extended corona would get in the way of any meaningful extrasolar imaging.  

    That is resolved easily enough by being farther away.  

    If instead of 550 AU, the ‘design distance’ was 1,000 AU, then (rather obviously) the apparent diameter of Sol becomes half as great, but the ‘best focus ring’ would also be closer to center. How much? ⅔ power. Which makes it ⅓ power further from Sol’s limb.  And so it goes. Is that enough to competently block the solar wind, extended corona and so on?  I still don’t think so.  

    More like 2,000 AU.  

    And then the view would be magnificent.  

    STILL, have the problem of “slowing down enough” to do 10, 15 years of observation.  

    ⋅-⋅-⋅ Just saying, ⋅-⋅-⋅
    ⋅-=≡ GoatGuy ✓ ≡=-⋅

  28. So, if we had 5.273 kg of eggs we could make 11.568 kg of bacon and eggs!
    Providing we had 6.295 kg of bacon.

    See! It’s done to 3 decimal places. That’s sciencey, that is.

  29. I’m very pessimistic on spacecraft utilizing megawatt level power.
    Take nasa’s proj prometheus (2003-2005), it specced out 450 m2(4840 ft2) of radiator to deal with 1mw of thermal output @500K.

    DOI: 10.2172/881290

    It’s not that it isn’t doable, it just adds a lot of metaphorical inertia to an already complex problem. This one is still in my “not going to happen in the foreseeable future” column.

  30. My (limited) understanding is that the gravity lens position is actually a fairly large region. It starts at 550 AU but extends out a fair bit further.
    Within that region you have to adjust your own local lenses and image processing as you keep drifting further out, but you can keep using it for quite a while.
    So you can travel to the start of the lens distance, and then start a period of observation that could last for years (depending on speed of course).

  31. Can some science evangelist please explain to me how can the gravity-lensing 550 AU mission be done without a deceleration phase..? Wouldn’t it be necessary to sorta stop to focus on something and then keep orbiting (a very long orbit) to take a spheric ultramegagigapic of anything that surrounds our system?

Comments are closed.