≡ Menu

Sunvoyager’s Pedigree: On the Growth of Interstellar Ideas

Kelvin Long’s new paper on the mission concept called Sunvoyager would deploy inertial confinement fusion, described in the last post, to drive a spacecraft to 1000 AU in less than four years. The number pulsates with possibilities: A craft like this would move at 325 AU per year, or roughly 1500 kilometers per second, ninety times the velocity of Voyager 1. This kind of capability, which Long thinks we may achieve late in this century, would open up all kinds of fast science missions to the outer planets, the Kuiper Belt, and even the inner Oort Cloud. And the conquest of inertial confinement methods would open the prospect for later, still faster missions to nearby stars.

Sunvoyager draws on the heritage of the Daedalus starship, that daring design conceived by British Interplanetary Society members in the 1970s, but as we saw last time, inertial confinement fusion (ICF) was likewise examined in a concept called Vista, and one of the pleasures of this kind of research for a scholarly sort like me is digging out the history of ideas, which in the Long paper I can trace through work in JBIS and the IEEE in the 1980s and 90s, where ICF was considered.

Vista itself appeared in the literature in the 1980s, drawing on this earlier and ongoing work, its conical shape a response to the potentially damaging neutron and x-ray flux that ICF produced. Long emulates its form factor in the Sunvoyager design. I should also mention a NASA concept called Discovery II that I hadn’t encountered until now, a spacecraft designed for a mission to the gas giants using a magnetic fusion engine. Both this and an early ICF design by Lawrence Livermore Laboratory’s Rod Hyde and colleagues in the 1970s would use an engine with a mass of 300 tons, a figure which Long selected for the calculations in his Sunvoyager paper as he validated the HeliosX code using Vista as the template: “The current level of accuracy will suffice for making predictions for the expected design performance of the Sunvoyager probe.”

So what do we get as we downselect to achieve the Sunvoyager design? The image below shows the concept.

Image: This is Figure 8 in the paper. Caption: Concept design layout of Sunvoyager spacecraft configuration. Credit: Kelvin Long.

Notice the radiators, a critical part of the design, for we need to find a way to reduce waste heat. Long notes that for Vista, the radiation interaction with the structure was about 3 percent – in other words, the vehicle intercepts about that amount of the neutron and x-ray flux from the fusion reactions. He assumes a higher figure for Sunvoyager, although adding that using a mixture of deuterium and helium-3 as the fuel (Vista used a capsule of deuterium and tritium) would reduce these effects. The design also includes an annular radiation shield within the engine structure.

Long assumes the use of X-band frequencies for communications, transmitting at 8.4 GHz with a power output of 100 W, the signals to be received via the Deep Space Network’s 70-meter dishes. It’s interesting that he does not push for laser methods here, wisely so, I think, given the pointing problems we’ve discussed recently at deep space distances. Pushing data back to Earth from 1000 AU is daunting enough:

The expected data rate at 1000 AU will be 1 kBits∕s. Backup medium- and low-gain antennas are also likely to be required. Note that radio signals from a distance of 1000 AU will take around 138 h to reach Earth receiving antennas, and so significant data latency should be expected. The high-gain antenna will be mounted on a rotatable fixing (rather than body mounted) and on a set of rigid extension poles so that it can always be pointed toward Earth, which avoids the need of having to rotate the entire spacecraft such as was performed for the Voyager 2 and New Horizons missions.

The Sunvoyager interstellar precursor probe would be assembled in Earth orbit following multiple launch missions. The author likens building the craft to the construction of the International Space Station, noting on the order of 10 launch vehicles may be needed to get all the parts into the assembly orbit. Booster rockets, perhaps nuclear thermal, would be used to move the vehicle away from Earth at 17 kilometers per second (which happens to be Voyager 1 speed). This reaches twice the mean Earth-Moon distance in a day or so, at which point the fusion engine can be ignited. And here we go with ICF fusion on our way to the outer Solar System:

A capsule is accelerated into the target chamber where the bank of laser beam lines can target it within the open reaction chamber to the point of thermonuclear ignition. A set of externally placed laser-focusing mirrors may be required to ensure a symmetric implosion. The plasma from the detonation will expand into the hemispherical target chamber, with the charge particles then directed by large magnetic fields internal to the chamber. These are then ejected for thrust generation while the next capsule is loaded onto the target ignition point. This occurs 10 times per second, although the hydrodynamic and nuclear phases of the ignition take place on microsecond and nanosecond time scales, respectively, so that in between each ignition there will still be around 10−5 s of time for the loading of the next capsule while the plasma from the previous one is being ejected.

The numbers on the ICF fusion for Sunvoyager are, shall we say, mind-boggling. Consider this: The mission needs 200 million fuel capsules, or 50 million per tank. This is, as the author comments, “no small undertaking,” a thought I can only echo. If we’re looking at constructing and flying a mission like this in, say, 50 years time, we may be able to assume advances in robotic automation and additive manufacturing, but we also have the problem of acquiring the needed fuel. You may recall that the Daedalus starship design was built around the notion of mining the gas giants for helium-3. That, in turn, assumes a Solar System infrastructure sufficient to make such mining feasible.

Image: This is the paper’s Figure 12. Caption: Concept design configuration (side view) of Sunvoyager spacecraft. Credit: Kelvin Long.

I like the sheer daring of concepts like Daedalus and Sunvoyager. Remember that when those frisky BIS engineers put Daedalus together, they worked at a time when it was largely considered impossible to reach another star by any means. Daedalus seemed impossible to build (it still does), but it violated no laws of physics and became a vast engineering problem. The point wasn’t that building it would bankrupt the planet. The point was that if we did decide to build it, nothing in physics would prevent it from working. Assuming, of course, that we did conquer ICF fusion for propulsion.

In other words (and Robert Forward would hammer this home again and again in talks and in papers), interstellar flight was not science fictional dreaming but a matter of reaching the appropriate level of engineering, which one day we might very well do. A mission design like Sunvoyager reminds us that we can stretch our thinking based on what we have today to make wise decisions about how and where we invest in the needed technologies. We gain scientific knowledge in doing this and we also rough out the roadmap that points to still further missions that one day reach another star.

Image: The extraordinary Robert Forward, wearing one of the trademark vests created by his wife Martha. Forward chose this photograph to appear on his own Web site.

So I think Kelvin Long is spot on in his assessment of what he does here:

Additional studies will be required to further develop the design configuration and specification for the Sunvoyager mission proposal so that it can be matured to the point of a credible mission in the coming decades to include a subsystem-level definition. However, the calculations presented in this paper show promise for what may be possible in the future provided that investments into ICF ignition physics are continued and then the applications of this technology pursued with vigor.

I think Bob Forward would have liked this paper. And because I haven’t quoted his famous lines (from JBIS in 1996) in their entirety since 2005, let me do so here. He’s looking into a future when we go from interstellar precursors into actual interstellar crossings to places like Proxima Centauri, and he sees the process:

Travel to the stars will be difficult and expensive. It will take decades of time, gigawatts of power, kilograms of energy and trillions of dollars. Recently, however, some new technologies have emerged and are under development for other purposes, that show promise of providing propulsion systems that will make interstellar travel feasible within the forseeable future — if the world community decides to direct its energies and resources in that direction. Make no mistake — interstellar travel will always be difficult and expensive, but it can no longer be considered impossible.

The paper is Long, “Sunvoyager: Interstellar Precursor Probe Mission Concept Driven by Inertial Confinement Fusion Propulsion,” Journal of Spacecraft and Rockets 2 January 2023 (full text).

{ 19 comments… add one }
  • Antonio January 11, 2023, 10:59

    “This reaches twice the mean Earth-Moon distance in a day or so, at which point the fusion engine can be ignited.”

    I find it difficult to believe that the rocket exhaust is so dangerous that the ship can’t be launched from LEO. Do you have some numbers supporting that? If 500 km of distance and a thick atmosphere aren’t enough protection, how could the ship possibly survive to that, even more, survive for years?

    • Kelvin F Long January 11, 2023, 17:58

      There are two issues:

      1. Intensity of the exhaust beam. Probably not a good idea to have any humans looking at it directly since it will be very bright. You could likely observe a launch from the ground the same way we might look for a comet today.

      2. Any particles associated with radiation, which is bad for life and bad for any nearby technology.

      i.e. you wouldn’t want to switch on the fusion engine near any Earth orbiting satellites or space habitats.

      • Antonio January 12, 2023, 3:53

        Sorry, still not convincing. I want to see some numbers. A point-like source (not a laser), 500 km above your head, behind a shield roughly equivalent to 10 m of water, and moving at hypersonic speeds to the east, away from you? It doesn’t seem dangerous at all.

        • Juraj January 16, 2023, 20:17

          It may have similar effect as Starfish nuclear test. It affected ground infrastructure and there were no satellites then.

      • Antonio January 12, 2023, 4:00

        As for satellites, if a satellite, say, 100 km away, can’t easily survive some minutes, how could the spacecraft survive, even with his shield, a year of rocket operation?

  • jlandau January 11, 2023, 11:08

    Whoops — you accidentally called it ‘SunDiver’ in two places. Reminds me of our good conversations on Sundiver missions all those years ago ;-)

    • Paul Gilster January 11, 2023, 11:08

      Yes, and Kelvin caught it himself. Now fixed. Old habits die hard…

  • Alex Tolley January 11, 2023, 14:07

    Just as Lubin provided intrasystem performance for his D-Star laser sails, so I am also interested in the intrasystem performance of this ICF-powered spacecraft. [Interestingly SunVoyager seems to have a similar cruise performance to Galaxy, the spaceliner in Clarke’s 2063: Odyssey Three.]

    Some rough calculations suggest that SunVoyager could operate under full power with a turnaround that would reach Jupiter in 36 days, and Saturn in 52 days.

    Such a large ship with short duration mission times, much shorter than current trips to Mars, suggests it should be crewed rather than robotic. With just a 0.03g acceleration, rotating crew quarters would seem a good idea.

    This puts me in mind of the Discovery in 2001:A Space Odyssey (although the Leonov in 2010 seems like a better solution for rotating crew quarters). [Galaxy had a far higher acceleration and needed no rotating quarters while under acceleration.]

    The Discovery is estimated to mass 5440 tonnes. This wet mass incorporating the 300-tonne SunVoyager fusion engine reduces the acceleration from 0.3 ms^-2 to 0.047 ms^-2. [Discovery had a lot more propellant tanks, but this is ignored for this BoE calculation.]

    Trip time to Jupiter is 66 days, and Saturn is 92 days, both well within the current Mars trip times.

    If such a ship can be built (wet mass the equivalent of a modern destroyer) requiring 50+ launches to deliver the components and probably a team to assemble and test it in orbit, it would transform our use of the solar system, for exploration and industrialization. Could it be built within the 21st century? IDK, but if the economy can be made to continue to grow, a deliberate effort to do so seems plausible. As with the various space stations that have been launched, there seems to be no need for fancy enclosed life support systems that recycle everything. Water recycling, freeze-dried food, and CO2 removal seem sufficient for life support. The propulsion provides the brute force means to use low-tech life support. [Could water be brought along as a supply of O2 for life support, and H2 to boost the thrust?] Only remote facilities in deep space that are difficult to resupply would need enclosed life support systems. It is a prospect with interesting ramifications for humanity’s future in space.

    • Geoffrey Hillend January 12, 2023, 18:57

      I like the size of the ship. My only concern is if the ICF uses a NIF concept, then it might have a high specific impulse, but very low thrust since one can’t fuse to many of those pellets at one time which is why I am skeptical of that design. With the size of a aircraft carrier one could have s interplanetary spacecraft which might have more than one type of thrust just in case one type of engine broke.

  • Kelvin F Long January 11, 2023, 16:19

    Since Discovery has come up several times in this and the last article by Paul, it’s probably worth me sharing that I do have an ICF (rendezvous) model of Discovery running but I have not yet published it anywhere since it’s a preliminary model and hence why I have not yet talked about it. So Centauri Dreams readers can be the first to read about it ahead of any paper. To recap:

    Discovery I: gas core nuclear fission Cavradyne propulsion system from the movie 2001 A Space Odyssey.

    Discovery II: spherical torus magnetic confinement fusion, this was a NASA study published in 2005.

    Discovery III: laser driven inertial confinement fusion design. This is one I have produced which attempts to run a mission to the gas giants but matching the masses and total trip time for the Discovery II mission. I run my model in the code HeliosX which I have been building.

    For the Jupiter mission to 4.7 AU the mission is completed in 118 days using 11 tons fusion fuel (DHe3) and 807 tons expellant augmentation which are injected into the exhaust stream to boost the mass flow rate. The total vehicle wet mass is 1,690 tons. Its crewed with 6-12 people. My models generate a thrust 28 kN, jet power 4.8 GW and specific power 15.1 kW/kg. The model currently assumes a 46 mg capsule mass and 24 Hz pulse frequency.

    For the Saturn mission to 9.57 AU the mission is completed in 212 days using 11.1 tons fusion fuel and 816 tons augmented expellant propellant. The total vehicle wet mass is 1,699 tons. The model generates a thrust 21 kN, jet power 4.8 GW, specific power 15.1 kW/kg. Again using 46 mg capsule mass and 13 Hz pulse frequency.
    Obviously, I can ramp up the pulse frequency or change the capsule mass and get the vehicle there sooner than the mission times stated above, but I was matching the missions of the NASA study. So Alex Tolley your estimates were reasonable although I would likely have to double the pulse frequency. At some point I will write this up into a paper, but I thought it was worth sharing that I do have a Discovery III model.

    Overall, I am constructing the roadmap to interstellar but on the assumption of ICF propulsion because that is my area of specialism. As an aside, since I also serve on the Breakthrough Starshot Advisory Committee, I have been looking at a similar roadmap for laser sail propulsion following the outstanding legacy work of Robert Forward and recent work by others such as Professor Phil Lubin.

    I should mention also that I consider myself a student of the likes of Robert Forward and Arthur C Clarke. Arthur in particular who I met had a big influence on me and I find his books inspiring to this day. I think bringing something like ‘Discovery’ to life should be a near-term goal for humanity…. and I hope its on the radar of Elon Musk because it really should be. Such a capability in space is a game changer for our species and has all sorts of ramifications.

    • Jeff Wright January 12, 2023, 4:47

      I also like Carlo Rubbia’s Americium 242 engine concept as a secondary means of propulsion…ICF is better as a drive-but NTRs better for power generation. Stan Borowski is another good contact-Steven Howe too.

    • charlie January 12, 2023, 16:02

      @Kelvin F Long
      “Both this and an early ICF design by Lawrence Livermore Laboratory’s Rod Hyde and colleagues in the 1970s would use an engine with a mass of 300 tons, a figure which Long selected for the calculations in his Sunvoyager paper as he validated the HeliosX code using Vista as the template:…”
      What is the HeliosX code ? Is it something that can be open sourced for review ?

      • Kelvin F Long January 12, 2023, 17:38

        Its a code I have written in Fortran 95 for running space propulsion problems in-line with a spacecraft mission profile and performance calculations. It was discussed in the acta astronautica paper cited by Paul. Its not currently available for other people to use since its still under development but at some point I may release it as open source for others to use and build on. Its around 5,000 lines of coding currently. I am also writing a manual in parallel with it that contains all the key equations and algorithms. You will appreciate I am still in the weeds with it but I have written it with the intention of potentially distributing it for the eventual use of others.

  • Ron S. January 11, 2023, 21:12

    “11.1 tons fusion fuel … and 13 Hz pulse frequency”

    Without meaning to be overly negative, when I see quantities like these I am pretty well convinced that drives of this capability will be built with propulsion technologies (or fundamentals) yet to be discovered.

  • Adam Crowl January 12, 2023, 16:18

    Hi Paul & Kelvin
    Interesting conceptual design, but it has some issues. The original conical design of the VISTA minimised the neutron exposure – a shadow shield protected the magnetic coil and everything behind it in its conical shadow. Looking down from above the VISTA vehicle would look like a conical pot with a hole in its base. Everything in the empty part of the “pot” would be directly exposed to the neutron shine of the fusion detonation point. Additional shielding would come from the thousands of tonnes of expellant carried in the walls of the pot, the outer part of which is lined with radiators for the lasers. Those are part of the laser assemblies, like Hyde’s original laser designs.

    So you can see why Kelvin’s modified version is somewhat at variance with the physical reasons embodied in the original VISTA concept.


  • torque_xtr January 12, 2023, 16:18

    There is the way for ICF propulsion that does not require technological breakthroughs and could be detectable from afar.

    A 0.3 Tesla magnetic nozzle several kilometers wide can contain and deflect explosion products of a 550 kt thermonuclear warhead. The best use of them if we knew that the Sun will explode in a hundred years, or we want to launch massive fast-moving probes =)

    While peak X-ray output of such explosion in space is comparable to stellar (1e20 watts), the peak flash is microseconds long and total integrated flux from even Alpha Centauri distance makes detection difficult (but not impossible). But there also will be very strong and distinct radiowave pulses from cyclotron radiation, which may be detectable from greater distances.

    The emission from electrons will have cyclotron frequency in several GHz, decreasing towards zero within milliseconds as fireball exits the nozzle through diminishing magnetic fields. Much more difficult to estimate but if 0.05% of the yield goes into radiowaves from 0.3 to 3 GHz within 10 milliseconds, then total RF power will be 1e14 W, and the peak flux from 10 pc would be 1e-22 W/m2 or 0.003 mJy. Ions would radiate at sub-MHz range with much weaker flux, but maybe comparable or even bigger spectral radiance. Not so out of reach, maybe some future observatory on the far side of the Moon, shielded and cooled to the temperatures of lunar night, will be able to catch that in some considerable volume.

    This is so much like FRBs that I can only regret that they come from distant galaxies and do not repeat regularly once in several seconds or minutes as nuclear blast ICF propulsion should do. (if ETs use such barbarian tech :-) )

  • Tampa Red January 16, 2023, 10:48

    It appears the biggest obstacle to subluminal interstellar travel will be the maintenance of earthside mission support over the long periods of time necessary for the spacecraft to reach its destination. Even at relativistic speeds, missions will be underway for decades, perhaps centuries, and each one will require a dedicated team back home to download data, upload commands, and otherwise manage the spacecraft.

    There is no fundamental technical reason this can’t be done, but I suspect maintaining bureaucratic support for these long-term flights will be difficult, if not impossible. Who will want to devote his career to working on a spacecraft launched decades ago and still decades from reaching its destination? This is especially the case when other missions, launched much later, perhaps with faster propulsion and more interesting targets, will be competing for talent and funding.

    • Paul Gilster January 16, 2023, 15:34

      On the other hand, the Voyager team persists after decades, so it can be done. But your point is a good one. Bob Forward wrote about the dilemma and consequences of mission support in his novel Rocheworld.

  • Mike Serfas January 17, 2023, 0:10

    I can’t shake the feeling that this design is something like an electric car powered with Leyden jars and electrostatic disc generators. I’m yet loath to give up my excitement from reports, now over two decades old and apparently not very repeatable, that “induced gamma emission” could release nuclear power on demand.

    Just recently I saw a paper to rekindle the old flame, though it made no claim in this direction, explained at https://phys.org/news/2023-01-entanglement-scientists-nuclei.html and published at https://www.science.org/doi/10.1126/sciadv.abq3903 . Somehow “quasi-real” 30 MeV photons manage to get from one gold or uranium nucleus to another during a near-collision, allowing a “probe” into the nucleus.

    I don’t pretend to understand half of it, and it rounds out the standard quota of a half dozen impossible things before breakfast with mentions of polarized gamma rays, femtometer imaging resolution, entanglement between non-identical particles, measurement of gluonic structure, and localization of specific bosons inside a nucleus. (Apparently gamma ray lenses are also a thing, and if they are weak, well, space is vast. https://www.wired.com/2012/05/gamma-ray-lens/ ) If there were some way to turn all that into some clever method to touch the nucleus and induce radioactive decay on demand… it might be a lot easier to build an interstellar vessel.

Leave a Comment