≡ Menu

Icarus: Revisiting the Daedalus Starship

by Kelvin Long

Project Daedalus was the first thoroughly detailed study of an interstellar vehicle, producing a report that has become legendary among interstellar researchers. But Daedalus wasn’t intended to be an end in itself. Tau Zero practitioner Kelvin Long here offers news of Project Icarus, a follow-up that will re-examine Daedalus in light of current technologies. A scientist in the plasma physics industry and an aerospace engineer, Long is assembling the team that will begin this work in 2010, following a ‘Daedalus After 30 Years’ symposium scheduled for September at the headquarters of the British Interplanetary Society. Can we improve Daedalus’ propulsion systems, change its targets, modify its shielding? Numerous theoretical studies await.

During the period 1973-1978 members of the British Interplanetary Society undertook a theoretical study of a flyby mission to Barnard’s star, some 5.9 light years away. This was Project Daedalus, which remains the most detailed study of an interstellar probe ever attempted. The 54,000 ton two-stage vehicle was to be powered by inertial confinement fusion using electron beams to compress deuterium/helium-3 fusion capsules to ignition.

Daedalus was to have obtained an eventual cruise velocity of 36,000 km/s or 12% of light speed from over 700 kN of thrust, burning at a specific impulse of 1 million seconds. Travel time to flyby at destination would be approximately 50 years.

Daedalus had three stated guidelines:

  • The spacecraft must use current or near-future technology.
  • The spacecraft must reach its destination within a human lifetime.
  • The spacecraft must be designed to allow for a variety of target stars.

Announcing Project Icarus

Project Icarus: Son of Daedalus – flying closer to another star. This is the full name of the new study, a Tau Zero Foundation initiative in collaboration with the British Interplanetary Society (BIS). Over three decades have passed since the Daedalus work, making this a good time to revisit the design study in light of scientific and technological advancements.

The purpose of Project Icarus is as follows:

  • To design a credible interstellar probe that is a concept design for a potential mission this century.
  • To allow a direct technology comparison with Daedalus and provide an assessment of the maturity of fusion based space propulsion for future precursor missions.
  • To generate greater interest in the real term prospects for interstellar precursor missions that are based on credible science.
  • To motivate a new generation of scientists to be interested in designing space missions that go beyond our solar system.

Certain reference points follow on from the original Daedalus study, modified to reflect changes of time and circumstance. Thus the spacecraft design must use current or near-future technology so that it could be credibly launched by 2050. It must be designed to reach its destination as quickly as possible, in a time-frame not exceeding sixty years but, hopefully, much sooner. The propulsion system must be mainly fusion-based. Assuming realistic maximum cruise speeds of 0.3 c and a sixty year flight duration, this places approximately forty-eight stars within an 18 light year distance within range of Icarus.

Genesis of a New Starship Study

In the introduction to the Daedalus study report Alan Bond states that:

“…it is hoped that these ‘cunningly wrought’ designs of Daedalus will be tested by modern day equivalents of Icarus, who will hopefully survive to suggest better methods and techniques which will work where those of Daedalus may fail, and that the results of this study will bring the day when mankind will reach out to the stars a step nearer.”

So in essence, the naming of the successor project as Icarus was suggested by the original study group.

Daedalus and Icarus were characters from ancient Greek mythology. In an attempt to escape the labyrinthine prison of King Minos, Icarus’ father Daedalus fashioned a pair of wings made of feathers and wax for both himself and his son. But soaring joyfully through the sky, Icarus flew too close to the sun, melting the wax on his wings. He fell into the sea and died after having ‘touched’ the sky. Project Icarus aims to ‘touch’ the stars and escape from the bounds of mother Earth.

Toward an Evolving Design

An assessment of the many dozens of propulsion concepts for interstellar flight made it clear that one way to advance the prospects for interstellar travel would be to focus on a specific design proposal. This way, a concept design could be derived, iterated and improved. Over time, the concept would be worked upon by future generations and could ultimately lead to a design blueprint for an interstellar probe. As the Daedalus study was performed three decades ago, it seemed appropriate to start by re-designing the Daedalus probe with updated scientific knowledge.

Thus the genesis of Project Icarus. It is hoped that other teams around the world will eventually be assembled, working on specific propulsion proposals that have been investigated in the past, such as Starwisp, VISTA, AIMStar or one of the many others. In this way, the technological maturity of different propulsion schemes can be improved over time, drawing on a common background of study rather than diverse and uncoordinated research efforts.

Needless to say, Centauri Dreams will provide regular updates on the progress of the Icarus design team, which will also be developing its own Web site in conjunction with the September symposium.

Comments on this entry are closed.

  • Didac March 19, 2009, 15:46

    An important point to keep in mind in this kind of interstellar probes is the research that can be made in the first months of years, studying interstellar medium (ISM). So, if a interstellar probe does not attain its “star target” in say 50 years is not so a big problem if some scientific results are continuously provided.

    It is very difficult to make predictions about the 21st Century. Some relevant changes will take place in demographics (a world zero-growth perhaps after the 2060s) and in energetics (diversification of sources), and of course in economics and politics (more difficult to predict because of the conflictive forces connatural to any modern human society).

    Anyway, the first truly “interstellar medium probes” were launched in the 1970s.

  • occam's comic March 19, 2009, 19:38

    Icarus can be completely different from Daedalus.
    First: weight let’s not talk thousands of tons for the weight but hundreds of grams. (the probe is manufactured from the nano scale up)

    Second: Don’t use exploding fusion bombs- Use a big electromagnetic launcher that remain in orbit around our sun ( that way you can launch many probe missions and they can be powered by that big fusion reactor in the sky, aka the sun)

    Third: The probe is made millions of meso sized parts (~100 cubic microns) that are launched separately and self-assemble on the way. (each part that is launched has computers, sensors, transmitters and way to maneuver in space.)

  • James M. Essig March 19, 2009, 20:35

    Hi Folks;

    The genesis of Project Icarus is truly inspirational. A final cruise velocity of 0.3 C seems very reasonable to me. Electrodynamic breaking mechanisms such as the deployment of a large highly conducting if not super conducting loops or wounds of cable could be used to build up tremendous currents thus producing a magnetic field that would act as a break when reacting against the interstellar magnetic field. Alternatively, a mag sail plasma bottle type breaking mechanism might be deployed to slow the craft down in a sort of reverse process anticipated for outgoing magsail craft.

    The ability of craft to cruise at 0.3C can open up the Entire Milky Way for human colonization. I like the idea of star hopping wherein humans would settle planetary systems around a given star and then launch other mission to adjacent stars and the process would repeat it self.

    Medical science to greatly lenghten human life expectancy would certainly be a help in promoting 0.3 C range generation ships.

    I have a tendancy to joke with my loved ones that I might actually be alive when the first interstellar mission is launched on its way to another star system. I will be 88 years old by 2050 and by God, we just might succeed in the above Project Icarus time table.

    It is truely a priveledge to think that this very century, we humans might end up being the ETs that land on an extra solar planet. And this might involve really nothing more than late 19th Century early 20th Century classical electrodynamics, special relativity, mid 20th century nuclear physics, and 21st century numerical flight control technologies. The giants like Einstein, Maxwell, Bohr, Fermi, Dirac, Shroedinger, Wheeler, and the like have laid the fundamental physics to permit such journeys a half century to 1 1/2 centuries ago.

    A bold international mission to send humans to the stars this very century would likely unit humanity as never before.



  • Adam March 19, 2009, 20:56

    Owww! Tough design requirements! Getting 0.3 c out of fusion rockets – which we haven’t yet built – is going to be tricky.

    Alan Bond, back in 1971, suggested a fusion rocket with a cruising speed of 0.15 c and a mass ratio of 22,000 (i.e. delta V is x10 exhaust velocity.) The design was meant to fly to another system within 10 ly within 60 years, based on best guess fusion-pulse efficiency just after Winterberg’s first public papers on electron-beam fusion ignition.

    Robert Bussard has suggested a top exhaust velocity of ~14,000 km/s for a Direct Fusion Product p+B11 fusion rocket. To get to 0.3c – and then deccelerate via a magnetic-sail or plasma-sail – the ship would need a mass-ratio of 760. Like Bond’s design such a mass-ratio would be achieved by massive amounts of staging.

    Alternatively we fire pellets of fusion fuel (boron-hydride?) along the ship’s acceleration track, which it then scoops and fuses on the fly. Accelerate fast enough and the track doesn’t have to be too long. For example at a constant 1 gee it’s 0.047 light-years long – about 3,000 AU. If the pellets are fired off at 0.01c then it takes a bit less than 5 years for the furthermost to be put in place. The real trick will be deccelerating the fuel pellet at capture and recovering the stored kinetic energy once it’s fused. How to couple the stored energy to the exhaust at high efficiency? I’m not altogether sure that’s a tractable problem.

    Thus perhaps we’ll have to push the ship using the pellets, though if zapped at a sufficiently high relative speed they might efficiently fuse and add energy. For example boron micro-sails might be boosted towards the ship via a laser and zapped via a proton-beam to cause fusion, the expanding plume of which is deflected via the ship’s magnetic fields. That approach might work, but I’d have to defer to the greater knowledge of the fusion experts. Perhaps pellets can be captured and fused?

  • Astronomy Link List March 20, 2009, 5:52

    This article has been added to the Astronomy Link List.

  • Administrator March 20, 2009, 9:18

    Adam wrote:

    Robert Bussard has suggested a top exhaust velocity of ~14,000 km/s for a Direct Fusion Product p+B11 fusion rocket. To get to 0.3c – and then decelerate via a magnetic-sail or plasma-sail – the ship would need a mass-ratio of 760. Like Bond’s design such a mass-ratio would be achieved by massive amounts of staging.

    Even so, that mass ratio seems low to me. But Icarus, at least, won’t have to worry about decelerating, being a flyby probe. Deceleration adds such huge layers of complexity to the project and stretches what looks like an all but impossible mission (given today’s level of proficiency at fusion) into the realm of the surreal. Thorny problems indeed, which is why it’s going to be so interesting to watch the design team kicking inertial confinement ideas around.

  • Administrator March 20, 2009, 9:22

    Jim wrote:

    It is truly a privilege to think that this very century, we humans might end up being the ETs that land on an extra solar planet.

    A wonderful thought, but do keep in mind that Project Icarus is designed as an unmanned probe, with no deceleration capability at destination. This is a follow-up of the original Daedalus concept — no humans on board — but we can all hope manned starflight comes somewhere down the road.

  • Goldstein Hovercraft March 20, 2009, 10:48

    What can a flyby probe accomplish that a suitably big telescope can’t accomplish?
    Of course, a flyby probe would be an engineering accomplishment, and quite thrilling as well, but I’m asking my question in the context of learning about extrasolar planets, rather than learning about propulsion and other technological challenges. It might be useful to do a price comparison, keeping in mind that for nearly all (but not all) large telescope designs, multiple stars could be viewed, so it might be worth comparing one telescope to a series of flyby probes.

    — Eric

  • Goldstein Hovercraft March 20, 2009, 11:01

    By the way, I didn’t mean to be cryptic when I suggested that not all telescopes could be used for multiple stars — I was thinking of using the sun as a gravitational lens, as described on this blog here: https://centauri-dreams.org/?p=785

  • James M. Essig March 20, 2009, 11:24

    Hi Adam and Paul;

    Thanks for offering your perspectives. If we only launch unmanned interstellar probes to other stars this century I would be thrilled.

    An interesting take on the fusion rocket concept that might make Project Icarus more feasible would entail some sort of parasitic system wherein the fusion rocket fuel tanks are composed of fusionable fuels. The fusion rockets would gradually be fed fuel scavanged from the fuel tanks thus permitting little mass waste. Such a system would add complexity to the mission and so might not be appropriate for our first interstellar flyby missions. Finding a suitable fusion fuel that can form strong enough tanks would pose a challange also.

    Note however that if we can develope fusion rockets that utilize their fuel tanks for fusion fuel, the prospect to going to the stars under nuclear power becomes easier and the Isp requirements of the fusion fuel are slightly less rigorous.



  • Peter March 20, 2009, 12:20

    According to the group at talk-polyell.org, IEC fusion is still not within our grasp, and may still be a longshot. I’d be interested in how project Daedalus/ Icarus compares to project Orion, given the 60+ year lead that nuclear fission technology has over fusion.

  • ljk March 20, 2009, 12:36

    This is very exciting news! I have been fascinated with Project
    Daedalus ever since it first came out. I have the original printed
    book BIS published in 1978 and I too wish it were online for all –
    maybe it is time this happened in light of facilitating Project Icarus.

    This being said, I have the following issues with Daedalus having
    three decades of technological advancements and new knowledge
    to fall back on.

    If we are looking at revising Daedalus, perhaps Icarus should be
    a REIMAGING of the entire concept. Daedalus is huge both in the
    size of the probe and the infrastructure that was required to make
    the vessel happen. This will need to change as much as possible
    if we want a real interstellar probe ready to go in our lifetimes.

    Mining helium-3 in Jupiter’s atmosphere to fuel the fusion engine
    and having to build Daedalus in orbit around its outermost Galilean
    moon Callisto are just two big examples of why Icarus won’t launch
    by 2050 if we go down the same road with this new probe.

    NASA has pushed the first manned Mars missions out to the 2040s,
    so I think we can forget even basic crewed expeditions to Jupiter
    before the year 2100 – unless another nation or the private space
    industry jumps ahead and takes the lead in this area.

    This is my next issue, the need for Icarus to be fusion powered.
    Just how far along is real, sustained fusion energy in terms of
    becoming a reality? And how much longer before we can place
    it aboard a spacecraft?

    I remember reading an article from a 1984 issue of Omni magazine
    where Lyle Wood said the fusion process for Daedalus envisioned
    by the BIS would actually burn up the probe before it could get
    anywhere. What has been done since then to rectify this situation?

    Will it be possible to discuss an Orion type star probe engine as an
    alternative? We can make this setup work by the 2050 date. Or
    will all the issues surrounding the idea of nuclear bombs in space,
    even for peaceful purposes, stop things before they can even start?

    All I know is that there are no other feasible propulsion systems that
    could get a probe to Alpha Centauri in a human lifetime or two which
    have even a hope of really being ready by 2050 besides Orion.
    Please prove me wrong here, however.

    Going back to Daedalus’ size: Thirty years later, do we still need a
    star vessel that is so darn big? I have been intrigued with the idea
    of needle probes for a while, which could spread out all over an
    alien solar system and return bits of data on different things to add
    up to one big whole picture.

    We could use a big “dumb” container ship to get all the needles to the
    designated star system: A vessel that only has to keep the real little
    probes inside safe for the long journey and of course provide the
    means to get them there.

    This might also resolve the issue I had brought up earlier elsewhere
    on Centauri Dreams about an AI system for an interstellar probe
    having to be actually intelligent and aware to deal with problems
    immediate to itself. I never did like the plan to just abandon
    Daedalus after it zips through the Barnard’s Star system, not only
    because it seems like such a waste of a still-functioning vessel but
    also if the AI aboard is truly aware, would it be equivalent to sending
    the intelligence on a suicide mission?

    Would the human team members be able to handle a star probe
    several light years away that suddenly decides it doesn’t want to
    explore Alpha Centauri just to make its primate creators happy?
    I bet they won’t be too happy to watch their decades of work give
    them the electronic equivalent of the finger.

    There are more issues, but this will do for now. To sum up what
    I have said here, we need to do a lot more with Icarus than just
    give it a more efficient engine and a faster computer brain if we
    ever want to at least see an interstellar probe leave the Sol system
    in our lifetimes, or at least our children’s.

  • george scaglione March 20, 2009, 12:58

    hello all there are so MANY good ideas above that i am taking the liberty of answering several people all in the same posting!! first of all…paul…great article,caught my imagination immediately! exactly the reasom this site exists is it not!? but as you say toward the end,or imply, yes i want a human crew somewhere down the line how in the world could i not.lol in some way manner shape or form i want to be a part of such a crew someday!!! didac yes sir e bob it is hard to predict the future of these projects! what shape and or direction they may take etc etc! but i do think that you make excellent points all round. occam i like what you said very much i too think that an electromagnetic launcher orbiting our star is something to be very much hoped for frankly i was never really crazy about the fusion bombs concept. jim, .3 C to get us all over the milky way – well,that fits right in with something i was saying yesterday.saw something on tv about how the sun will absorb earth in about 5 billion years! but I say in 5 billion years we will be either all over the galaxy or extinct!!!! obviously we all see how your point fits in.good thoughts. adam,last and faaar from least – accurate way of looking at it – kinda hard to work with fusion rockets which we do not yet posses! very respectfully to all my friends above and with hopes of hearing alot more on this great subject.your friend george

  • James M. Essig March 20, 2009, 13:56

    Hi George and ljk;

    Thanks very much for sharing your persective.

    In terms of fusion rockets, every little bit of additional Isp that we can extract from the fuel is critical when planning future fusion powered unmanned interstellar probe flybys and manned star ships. I like the mantra of using antimatter to induce nuclear fusion.

    If the hydrogen is fused in batch mode, perhaps fusion can be effected by using small quantities of antimatter which may be carried on board from the start of the mission and/or produced in route.

    In an alternative scheme, the reactants produced by the fusion reaction could be saved if some sort of confined fusion reaction process is used to power an ion, electron, or photon rocket propulsion system. The fusion by products could be collected, cooled, and then recylced in fusion reactions involving the initial nuclear reaction products to form heavier elements yet, and the process could continue until the most stable form of iron is reached. If there is any antimatter left, the antimatter especially if the antimatter is in the form of antiprotons, could be made to interact with the iron nuclei thereby causing an additional energy release.

    The fusion reaction series could be choosen so as to extract the greatest quantity of energy from the series of nuclear reactions. Such a system would no doubt be complicated, but it offers an intermediate step between pure low atomic mass fusion reactions and matter antimatter reaction powered space craft. Isp is a precious commodity in nuclear powered star ships and must be optimized.



  • Goldstein Hovercraft March 20, 2009, 13:57

    A follow-up regarding telescopes vs flyby probes: I found this discussion at unmannedspaceflight.com: http://www.unmannedspaceflight.com/index.php?showtopic=5440 The discussion evolves into a somewhat off-topic conversation about artificial intelligence, but have a look at post #3 and post #5. The author makes the point that a terrestrial world might be – cloudy – on the day a flyby probe makes its closest approach! Either a telescope back here in Earth’s neighborhood or an interstellar probe which goes into orbit around another star would have much better luck of geting a clear view! Finally, there is there are two questions lurking in my query: how close to the most interesting extrasolar world will the flyby probe get, and how good are the telescopes and other instruments onboard?

  • Brian Wang March 20, 2009, 14:02

    If it is a fusion rocket, there was the idea that I and the mini-mag Orion people had mentioned of something else (another slower rocket or mechanism for accelerating fuel pellets) providing in -flight refueling.

    The fuel trail is laid out or if lorentz force acceleration or magnetic acceleration of fuel pellets and just run those into a pusher. Then you can get a mega-scale ship up to high fractions of light speed. As Adam talked about deliver the fuel. With a tough pusher plate (ala Orion) we do not have to slow the fuel up. We just let it hit so that it fuses on impact. Have ablation to protect the pusher plate.

    going nano-scale with as occam’s comic is talking is introducing probable but undeveloped technologies.

    If targeting of fuel pellets or beams over a distance is a problem then the design needs to have higher acceleration so we are done before it is too far away. Then the problem is providing the medium so that people can live through the high acceleration phase.

    Deceleration using magsails. Or pre-launch a slower ship that has the deceleration mechanism.

    Staged designs can reduce the constraints and make higher performance designs more achievable.

  • Administrator March 20, 2009, 14:10

    Larry wrote:

    If we are looking at revising Daedalus, perhaps Icarus should be
    a REIMAGING of the entire concept. Daedalus is huge both in the
    size of the probe and the infrastructure that was required to make
    the vessel happen. This will need to change as much as possible
    if we want a real interstellar probe ready to go in our lifetimes.

    Good point, and I think the question of how closely to adhere to the original design will be a major one, given issues like the helium-3 collection and others that you mention. I do believe that’s part of the exercise, though, to look at the earlier parameters and see where they can be adjusted and where completely new models need to be applied. Like you, I find a 2050 departure date vastly optimistic, given that our next, obviously unmanned, Jupiter system satellite won’t get there until the late 2020s. I think we can consider these ambitions rather than hard design goals, however. Massaging the concepts should provoke a lot of good ideas.

  • Administrator March 20, 2009, 14:13

    Goldstein Hovercraft wrote:

    Finally, there is there are two questions lurking in my query: how close to the most interesting extrasolar world will the flyby probe get, and how good are the telescopes and other instruments onboard?

    I suppose proximity to extrasolar planets will depend on the star chosen and what we learn about its system before any launch. Like you, I think the concept of a flyby probe was more viable in the Daedalus days, when we weren’t making such rapid strides in exoplanet imaging. But to repeat what I said to Larry, part of this process is in asking questions like this to figure out just where we are in our thinking on these possible spacecraft, and how much the underlying assumptions may have changed.

    People sometimes ask, for example, about the choice of Barnard’s Star for Daedalus. Back then, it was the only star with evidence for a planetary system (evidence which has turned out to be insufficient, although we still don’t have a definitive read on whether there are planets there). So one part of reimagining Daedalus is to drastically re-do its targets!

  • Administrator March 20, 2009, 14:17

    Brian Wang is talking about something similar to Jordin Kare’s idea of a ‘fusion runway,’ I think. Kare also developed an idea for the ‘Bussard buzz bomb,’ which would accelerate along a line of pre-deployed fusion pellets and use them for fuel — this one was a hollow ring of a spacecraft, like a big Cheerio.

  • ljk March 20, 2009, 14:32

    In 1980, Robert Freitas came up with a variation on the interstellar
    probe design, one that would reproduce itself at the target star
    system to make more probes to explore more star systems,
    reproducing at each new system.


  • Adam March 20, 2009, 20:53

    Larry’s linking to Freitas’s old paper is handy because it contains a detailed comparison of subsystem masses of both “Daedalus” and “REPRO”, conveniently tabulated – you have to glean the masses across several papers in the “Daedalus” JBIS issue. Freitas unnecessarily encumbered REPRO by assuming a 0.12c cruise speed, even though REPRO then took 1000 years to reproduce. Half the assumed speed would’ve allowed a 50-fold mass reduction.

    Several points for improvement of “Daedalus”…

    “Daedalus” needed a central computer because of the expected complexity of the AI and the pessimistic computer masses assumed. But computers of the required intelligence will be much smaller by the time “Icarus” launches.

    If we fire off the “probe” via an energy beam (pellets, plasma or neutral ions) then minimising size to reduce required power and maximise acceleration is a must – so why not launch the “probe” as a whole bunch of small independent, but intercommunicating units? With fold-out optic technology and laser communications there’s no need for the main-engine bell to double as an antenna, and because the sub-units can be launched independently there’s no need for a main vehicle to maneuver over several AUs to place sub-probes like in the original design. We can assume telescopes back in the Solar System will have placed all the planets in the target system, so maneuvering can be minimised. Smaller probelets will also have lower collisional cross-sections reducing the collision mitigation requirements substantially.

    One critical issue will be power. RTGs won’t be suitable because of their exponential decline. The recently announced “travelling wave” fission reactor might be suitable since it has an expectedly high burn-up fraction. A purely solid-state power converter might also be available – for example the thermal converter from the guy who invented the Super-Soaker. Thermal photovoltaics might also be available at high efficiency, or high-efficiency thermoelectrics.

    Another area for improvement is superconductors. “Daedalus” used liquid helium chilled aluminium. We can do better and by 2050 we might even have “room-temperature” superconductors available for magnetic-sails or plasma-magnet sails.

    “Daedalus” also assumed heavy capacitors for powering the electron-gun system. If the vehicles are launched via stay-at-home beam-systems then the whole elaborate gun-system can be discarded. What maneuvering is needed can be done via high efficiency Helicon plasma thrusters, or even simplified VASIMR technology. Station-keeping, especially for interferometric studies while in cruise, can be done via the photon-thrusters we’ve discussed here before. And, unlike “Daedalus”, we can assume high-efficiency solid-state lasers will probably be available at 0.1-10 MW power-levels. Perhaps high-reflection photon-thrusters can allow the sub-units to dodge interstellar dust without expenditure of propellant mass?

    So, some ideas for discussion. “Daedalus’s” design was quite detailed and I’d love to see more updating from more knowledgeable TauZero practitioners. I suspect the mass could be cut back dramatically even assuming fusion pulse propulsion.

  • Administrator March 21, 2009, 8:50

    Useful and detailed material, Adam. I think you’re exactly right about the possibility of dramatic mass reduction. Of course, we could always follow Robert Freitas’ lead when he turned REPRO into the ‘needle’ probe idea empowered by nanotech. Talk about mass reduction!

    All the comments posted re Icarus will be seen, of course, by the design team, and we’ll be tracking their preliminary work closely here. Keep the ideas flowing… As I mentioned, the initial symposium to kick off the formal project is in September, and it will be interesting to see what the time-frame is on Icarus vs. Daedalus — I would anticipate Icarus will be a five-seven year study, but we’ll see.

  • James M. Essig March 21, 2009, 12:10

    Hi Adam;

    Great comments by the way.

    Regarding using pellets to fire off the probe, I remember reading about the concept of antiproton calalyzed fission bomblets wherein molecular cages containing antiprotons within pellets of U-235, and if my memory serves me correctly, also combinations of U-235 and U-238 ,would be opened up thus causing the anti-protons to be released thereby causing a temporary rate of fissions to procede at a rate in which an effectively super-critical mass would be produced.

    If such atomic fission bomblets can be produced, perhaps they can also be encased in the best performing nuclear fusion fuels for the purpose of propelling a space craft to relativistic velocities. It might be easier to use such a fission fusion mechanism than it is to use direct anti-proton based fusion.

    Also, I have recently read a report regarding the U.S. military’s research on directed energy weapons, more specifically on 0.1 MW to 1 MW class lasers. The U.S. Army’s High Energy Tactical Laser is, I believe, an efficient solid state laser in the 0.1 MW to 1.0 MW range. Such laser technology, if it could be made to run continuously without burning up would potentially be excellent for beam driven sails.



  • ljk March 21, 2009, 13:14

    We should also consider what kind of information messages we
    need to put aboard Icarus for either future human explorers or
    ETI, or both.

    Especially if the probe does end up drifting through the galaxy

    And have we considered how the inhabitants of another solar
    system might feel about seeing something unknown come at
    them at relativistic speeds? One way to wipe out an entire
    Earth-type planet of its life is to hit it with just one interstellar
    vessel going at a good fraction of the speed of light.

  • James M. Essig March 21, 2009, 14:04

    Hi ljk;

    Those are important considerations.

    Even a 1,000 metric ton rest mass vessel traveling at 0.86 C or with a gamma factor of two would have the kinetic energy of the equivalent of 1,000 metric tons of matter converted into pure energy.

    This would be the equivalent of about 22.5 Teratons of TNT or 22,500,000 Megatons. This is the equivalent of about a one trillion metric ton asteriod hitting the planet at about 14 kilometers per second. A one trillion metric ton silicate asteriod is about 5 miles in diameter and a one trillion ton hard iron mickel asteriod is about 4 miles in diameter. These sizes are of the same order of magnitude as the asteriod that totaled the dynasours and which may have blasted a crater 175 kilometers in diameter all the way down to the Earth’s upper mantel.

    Perhaps any such probe could be designed with a self destruction mechanism such as a nuclear device which would completely vaporize the craft once its effective mission was completed.



  • Peter March 21, 2009, 14:29

    Will it be possible to discuss an Orion type star probe engine as an
    alternative? We can make this setup work by the 2050 date. Or
    will all the issues surrounding the idea of nuclear bombs in space,
    even for peaceful purposes, stop things before they can even start?

    That’s exactly what happened: public fear and ignorance about nuclear power coupled with fear of what the army might do with an orbiting nuclear warship:


    Like I said, nuclear fusion and antimatter (and of course ‘solar sails) is rather redundant, since nuclear fission will get us anywhere we need to go in our solar system, and it is technology that is well into the engineering (rather than the theoretical or experimental) stage. An antimatter rocket might get us to Pluto in one week instead of four months, but does that really matter? Neither fusion or antimatter are practical for interstellar missions, so we might as well go with what we know.

  • ljk March 21, 2009, 14:58

    James Essig said:

    “Perhaps any such probe could be designed with a self destruction mechanism such as a nuclear device which would completely vaporize the craft once its effective mission was completed.”

    I can see a number of issues with that idea, among them the one
    I brought up earlier in this thread about the AI aboard the starship
    being intelligent and therefore does that make it a being with rights.

    An “aware” intelligence on Icarus might not be very thrilled to
    know it has to commit suicide after a successful space mission
    as its “reward”.

    Being pragmatic, once the probe has gone through a designated
    star system, how does destroying it after the fact lessen the
    potential concerns of any ETI in that system who would probably
    be much more concerned about the ship when it is first arriving
    at their celestial home in an unknown state moving at relativistic

    I can see no sensible reason for a self-destruct mechanism on
    Icarus. If another species considered it a possible threat and
    they were advanced enough, they would likely destroy it on
    their own terms.

    Well, here we are, the poor vessel hasn’t even gotten on the
    proverbial drawing boards and we are already contemplating
    its demise! Seriously, I do hope we will think about what will
    become of Icarus once its main mission is done. I never liked
    how the BIS just assumed Daedalus would fly off into the galaxy
    and that would be the end of things.

  • Administrator March 21, 2009, 15:23

    Re fission vs. fusion, I am not on the design team and don’t know how much of a clean slate they are willing to make, or whether the design will be evolutionary as part of a continuing process that could refine the original model. It’s an excellent question, and I’ll pass along anything I learn about it. In any case, it would be interesting to see an Orion-style vehicle designed in the same way Daedalus/Icarus are being done.

  • Adam March 21, 2009, 22:53

    “Orion” as an interplanetary vehicle used nuclear fission reactions, though probably boosted by a bit of tritium in the centre for more neutrons. Dyson’s interstellar version used fission-triggered fusion bombs – still the only proven way to get net fusion energy. The performance he assumed (0.03 c) meant a very high fusion burn-up fraction, which may be hard to guarantee. But at least we know it can be done, albeit for very large scale vehicles (100,000 tons or so.)

    Bob Zubrin’s Nuclear Salt-water rocket uses greatly enriched dissolved uranium oxide which he proposes can get an exhaust velocity of 4,700 km/s. Let’s assume a 60 year flight to Alpha Centauri – thus a d-V of 0.07333c and a mass-ratio of ~108. For 0.1c it needs a mass ratio of 422. Pretty marginal performance.

    I know James has written elsewhere about really big mass ratios, but building such gargantuan vehicles can be ruled out by 2050. And I’m not optimistic that fission or fusion pulse drives can be minaturised effectively for 1 kg payloads with 10,000 or 100,000 mass-ratios.

    So while we know we can build vehicles pushed at decent efficiency by megaton-plus H-bombs, I don’t think that’s a reasonable goal for 2050. If we were moving large industrial facilities to Lalande 21185 or Sirius…

  • Adam March 22, 2009, 0:42

    A handy summary of fusion reactions can be found here…

    Bussard Fusion Systems

    …which mentions exotica like pure Helium-3 fusion or Lithium-6. The last has the advantage of being a metal of reasonable strength in the expected temperature range, thus easily converted into tankless fuel spheres or cylinders. The rest of the associated web-pages discusses several fusion starship designs, one illustration of which is mostly a very large cylinder of isotopically pure lithium-6…

    Explorer Class Starship

    …it masses 25,000,000 tons, thus requiring about 337 million tons of lithium to be mined and processed. But the crew vehicle itself masses over 100,000 tons, thus the vast scale. If we could launch a decent payload for 1/10th of “Daedalus’s” assumed 450 tons, then we might be able to launch a lithium fusion starship to 0.3 c massing at light-up some 16,300 tons, assuming the vehicle “dry” mass is twice the 45 ton payload. That assumes perfect fusion and total burn-up.

    So is it doable? A 20,000 ton starprobe is quite an undertaking, but the assumed dry mass is just 90 tons. That’s a single launch via an Ares V. The rest can hopefully be found in-space. About 400,000 tons of lithium needs to be processed, preferrably from asteroidal materials. The final ship would probably be a cylinder or sphere of lithium, dwarfing the engine and payload. The real problem area will be achieving a high power-to-mass ratio and high-efficiency fusion of a relatively difficult fusion material.

    Pure Helium-3 or deuterium-helium3 fusion might seem like good options based on the figures in the tables linked to, but they have several problems. Firstly Helium-3 is damned difficult to store as it can’t be frozen at low pressure, and once frozen it’s a super-solid, thus rather liquid like anyway. Second the D-He3 reaction isn’t as aneutronic as it’s advertised, which Larry has already mentioned. Side reactions of D+D produce copious neutrons and that’ll need heat-management in whatever neutron absorber is used – ironically lithium is the absorber of choice in current tokamak designs, and is vital to breed the tritium the reactions use.

    So lithium-6 does seem a good choice – it’s just hard to fuse.

    Fuel pellets striking a pusher might work. Can an ablatively cooled target be designed to take fusion-inducing impacts? Remember the “Orion” pusher-plate only interacts with plasma directly, not a compact impactor. Perhaps a micro-sail of lithium can be blasted into a plasma just before impact? I love Brian’s idea of accelerating the pusher pellets in Jupiter’s magnetosphere but I’d want to see a feasibility analysis. Is anyone here an electromagnetics expert?

  • James M. Essig March 22, 2009, 10:16

    Hi ljk;

    Thanks for offering your insights.

    An AI probe with its own mind would be a terrible thing to waste, and I would not want to have a conscious AI probe terminate its own existence. Our understanding of neuroscience, and psychophysics is advancing by leaps and bounds and so we humans might end up being the creators of a new type of conscious sentient lifeform.

    Perhaps a probe with optimized Isp from nuclear fusion power sources could reasonalby be launched for a terminal cruise velocity of 0.15 C to 0.2C and then use reverse thrust to slow down at the target star system. We would need to scape every bit of Isp from the rocket fuel used.

    Perhaps a huge hydrogen collector augmented perhaps with a magscoup type of apparatus could collect enough fuel for the final craft slow down and stellar orbit insertion.

    I like the scoop idea because of its drag inducing potential coupled with the fuel collection mechanism.



  • george scaglione March 22, 2009, 11:59

    ljk/jim…very good points above yes we need to be real careful ! however in my humble opinion the possibility of actually doing such damage is probably very slight to say the least.hahaha the “aliens” might be also, so advanced they would not see such a thing as a problem! “shields up!” – does that sound familiar!? thank you george

  • James M. Essig March 23, 2009, 13:42

    Hi Adam and George;

    Adam, I am not an expert on electromagnetic theory, however I thought that I would mention the concept of using small light sails, or mass driver driven solid and rigid pellets in conjunction with an incomming pellet fusion stream relative to the flight path of a relativistic fusion rocket star ship.

    If the outward bound light sails and the in bound fusion runway pellets are timed to form a collision within a ship based reation chamber or energy collection chamber wherein, with respect to the ship, the center of mass frame is used, the collisional explosion will be symmetrically and evenly distributed in terms of backward and foward direction radiation products with respect to the ship.

    If it is desirable to have the decay products skewed in either the forward or backward direction as might be preferable depending on the kinematical operation of the blast energy extraction mechanism, the the lab frame with respect to the relativistic star ship could be utilized.

    Either way, after reading your comments about mass beam driven sails, I have come to really like the concept.

    If a star ship was traveling at a gamma factor of 1.05 and the outward bound sails or pellets were traveling at a gamma factor of 1.05 with respect to the ship center of mass frame, then the appearent collision relativistic kinetic energy within the ship CM frame energy extraction chamber should be equal to 0.10 times the rest mass of either the fusion pellet or the outward bound pellet provided they have the same rest mass.

    If such a star ship some how could be accelerated to a whopping gamma factor of 1,000 using 1 gram outward bound sails or pellets colliding with inward bound fusion runway pellets, or in this case possibly inert matter pellets, then the blast energy in the CM frame would be equal to the equavelent of either mass traveling at a gamma factor of 2,000 with respect to the ship and the collision energy would be symmetrically released in a symmetric pattern in the amount of the equivalent of 2 kilograms of mass converted into energy. While I am not expecting gamma factor 1,000 star ships moving translationally through space any time soon, the first case might be doable this very century.



  • david lewis March 23, 2009, 15:24

    A 20,000 ton starprobe is quite an undertaking, but the assumed dry mass is just 90 tons. That’s a single launch via an Ares V.

    If the nuclear cannon mentioned on the Next Big Future website a while back is doable then even the 20,000 tons could be launched from earth. I vaguely remember the estimated price for a single launch of such a cannon as being around 1 billion dollars.

    Rather than being used for star probes such a vehicle could also be used to carry multiple probes to the sun’s gravitational lens. Though for such a shot distance one would probably not need the whole 20,000 tons – a speed of 0.001C would get you to the lens in a single year. One could imagine a vehicle that gets up to a decent speed and then launches 9 ten ton probes, each one aimed for a different point on the sun’s gravitational lens so that it targets a different star.

    Ten such vessels, which could be done with a single use of such a nuclear cannon, could allow for 90 such probes. One could then really start to examine the closest 90 star systems, the ones we are most likely to launch any interstellar vessels to, while at the same time getting a good grasp of just what interstellar space is like in terms of dust and other dangers.

  • Adam March 24, 2009, 6:02

    Hmmm… Launching 20,000 ton starprobes via nuclear cannons has such a deliciously scandalous feel in this age of enviro-Puritanism.

    Karl Schroeder has discussed Brian’s Verne Gun (as Karl dubs it) over on his blog and I must say it is such a cool idea, but also so utterly outrageous that it’ll remain an idea unless we have an utterly compelling cause to launch battleship masses into orbit. SSTO nuclear pulse starships might be such a cause if we were evacuating the planet.

    But I have such a dreadful feeling that “Orion” and Verne Guns will remain just ‘good’ ideas. But wouldn’t it be such an awesome sight, the launch of essentially an O’Neill Island One directly into space? Or a fleet of such, to claim the Solar System for all Earth-kind?

    For now our starships have to be less outrageous. Karl’s Twitter posts are discussing launching 1440 tons a day via laser – enough for a 1 GW, 3,000 ton SPS to be launched every couple of days. At that rate we’d have over a terawatt per decade in space. Enough to launch Sail-Beam propelled missions…

  • Howard T. March 24, 2009, 8:11

    I wonder if it would even be possible to get radio or laser telemetry/results back from a probe in the close vicinity of another star?

    If it would be desired to have a computer controlled ship destructed for some reason, you’d need merely to program it to flip a switch at the appropriate time. No need to tell it what it’s for. You can tell I don’t buy the concept of an artificial intelligence being developed, put in charge of a starship or being a ‘real’ being…. sorry….

  • Administrator March 24, 2009, 8:15

    Some years back I interviewed JPL’s James Lesh about the communications question. Having made an in-depth study of the matter, Lesh is confident that a probe in Centauri space could communicate its data to Earth via optical methods.

    A key paper for those interested: James Lesh, C. J. Ruggier, and R. J. Cesarone, “Space Communications Technologies for Interstellar Missions,” Journal of the British Interplanetary Society 49 (1996): 7–14. Also be aware of Alex Harwit, Martin Harwit, and Joss Bland-Hawthorn, “Laser Telemetry from Space,” Science 297, no. 5581 (July 26, 2002): 523, where the theoretical case for lasers is developed.

  • ljk March 24, 2009, 10:12

    Perhaps Icarus should also carry some capsules with data storage
    units to return information to Earth about the target star system in
    the even that the radio or laser communications system somehow
    fails while the probe is collecting data on the star system.

    I cannot imagine a much bigger tragedy for our first interstellar
    space mission than being unable to communicate with Icarus
    and never receiving all that precious information about another
    star system after waiting fifty years or so to get there.

    Having a means to ensure that the priceless data gets back to
    humanity is vital to the success of this mission. I know there
    will be a number of issues with a return data capsule such as
    the requirement of its own propulsion system, but even if it
    takes as much time to get back as Icarus took to get to that
    star, it will be better than never receiving it at all.

    I also do not want to assume that something better and faster
    will come along by then. We do not know what could happen
    that might keep another interstellar mission from being built
    and launched, so we must assume Icarus could be it for a long
    while and that its success is paramount.

    The data return capsules could also serve as backup storage
    devices for the Icarus computers, or if the probe is somehow
    damaged during the mission. The probe could be automatically
    programmed to eject the capsules and return them immediately
    to Earth in the event of any problem that would render the
    vessel unable to perform its duties.

    The capsules would be designed to survive even a catastrophic
    explosion, just like a black box data recorder on airplanes.

    Even if the capsules are unable to leave Icarus or the target
    star system, they would serve to preserve the records and
    data of the mission until another probe could arrive to assess
    the situation and make up for the first probe’s loss. Yes, I
    know I said don’t assume another interstellar mission will
    happen right after Icarus in case something goes wrong, but
    you know what I mean.

    The data capsules will have invaluable information on the
    star system as things and events were during the mission,
    which will not be possible to repeat even with a faster followup
    mission. Even if no one finds them for a very long time, they
    will then serve as important time capsules to the future, or
    as a message and evidence to any ETI that might find them.

    They will be a bonus to the information packet following in the
    footsteps of Pioneer and Voyager that should also be on Icarus
    and any other space vessel sent into the galaxy.

    Howard T., regarding your lack of confidence in an AI with actual
    awareness and intelligence – what hard evidence do you have for
    this? At the very least, Icarus’ electronic brain has to at least
    ACT like it is aware and intelligent to be able to handle any issues
    or sudden new discoveries during the mission, with help being
    litereally years away. Perhaps the AI’s programming will be so
    sophisticated that we will not be able to distinguish between
    artificial programmed intelligence and real consciousness.

    We shall see what advances in technology and our understanding
    of how the mind works will do for AI in the next fifty years.

  • Howard T. March 25, 2009, 6:37

    Blythly saying we should send back data capsules is one thing. Doing it is another. They would have to be decelarated and then re-accelarated back toward us. They’d need beacons of some sort, so we’d know they are coming. It’s a great idea, but it would probably multiply the size of the probe.

    But my question really is directed to whether even a focused radio beam could maintain communications at interstellar distance? This is probably a multiplication of the aiming and pointing problem we haven’t really got operating yet on space telescopes. The signal would have to be separated from the stellar radiation, as the probe will be quite close to, if not directly in-line with the star. I’m no radio expert, but I don’t think we can do it now? Can we do it in the future? Adjusting the beam would require either reaction mass or energy. And both would be hard to come by in interstellar space, if they weren’t on board.

  • Administrator March 25, 2009, 8:23

    Howard T. says:

    But my question really is directed to whether even a focused radio beam could maintain communications at interstellar distance?

    The paper by Lesh and Cesarone that I cited above points out that lasers are the medium of choice for interstellar data return, and goes into considerable detail in describing how it could be done. Radio offers nowhere near the efficiency and is deeply problematic at Centauri-type distances. JPL has done extensive work, much of it under Lesh, on refining optical communication techniques for future space missions as we move up the frequency ladder.

  • ljk March 25, 2009, 10:51

    I was not trying to be “blyth” about the idea of data return capsules.
    I put the idea out there to get things started on the subject and I
    did mention several issues with it.

    We better have some kind of backup plan to store and preserve
    the data that Icarus will be gathering from the star system it is
    sent to explore, otherwise this will be a very expensive engineering
    test, assuming we do not lose the engineering data in the process
    as well.

  • Howard T. March 26, 2009, 7:45

    I’ll study the papers you referenced. Thanks very much.

    But my instincts tell me that separating the laser light from the light of the star behind it (from our point of view) would be very challenging.

    Re designing an AI pilot: That would require the designers to guess what might be available when or if Icarus is launched. And to do so within two years from the end of the design conference. Roughly Jan. 2010? That’s daunting! But we already know how to design,build, launch and update a control program for an unmanned probe. We’re doing so on the Pluto probe. There are no doubt other contingencies it will have to handle, both forseen and unforseen. But if we are in communication with it and get data back that indicate problems, we can make some decisions here and send it instructions and updated control programs….. up to some point.

    Perhaps we should contemplate getting it slowed down and in orbit about the star or a planet? After a period of data gathering, it could accelerate away and return or continue to another star. Or far away enough to increase the chance of reliable communucation.

    I imagine this would require magnetic refueling of hydrogen and a really huge power supply.. But, what the heck… dream big.

  • Howard T. March 26, 2009, 8:18

    The article on the James Web Telescope is interesting. But it mentions the signal to noise ratio being near one for disturbance of the light….

    Somewhere I read a post about the problems of pointing it with sufficient accuracy. As yet unsolved, or a least unproven…

    That’s the same problem as shining a laser back at earth I would think? And would a laser disperse?

    Would aero-braking around a destination planet would be within the scope of the probe? I imagine we’d have some knowledge about it in advance.

  • Administrator March 26, 2009, 8:33

    Laser dispersion is far, far less than radio. The beam is tightly collimated and Lesh and Cesarone believe the signal would be readily discernible.

  • Administrator March 26, 2009, 8:41

    Howard T. writes:

    Perhaps we should contemplate getting it slowed down and in orbit about the star or a planet? After a period of data gathering, it could accelerate away and return or continue to another star. Or far away enough to increase the chance of reliable communucation.

    It would be wonderful to decelerate in the destination system but utterly beyond the range of a mission like this. Icarus (like Daedalus) uses vast amounts of fuel to reach cruising speed. Pushing the mass needed for the deceleration propellant runs the mass ratio off the scale. In other words, you would have to initially accelerate both the propellant for the initial, lengthy burn and all the propellant needed for the deceleration upon arrival.

    Matloff and Mallove provide the best explanation I’ve seen in their Starflight Handbook, though I also tackle this in my Centauri Dreams book, if nowhere near as well. Adding fuel to push still more fuel quickly makes the equations go crazy. Marc Millis once calculated this in terms of ion propulsion. A next generation ion engine would need more than 500 propellant tanks the size of supertankers to complete an Alpha Centauri flyby within a century. If you want to decelerate the same payload in the Centauri system, those five hundred supertankers would need to be supplemented by another three hundred million supertankers to make the 100-year journey and stop.

    The mass ratio problem is why so many of us are interested in beamed propulsion concepts, leaving the propellant to be supplied from the Solar System via laser or microwave or, potentially, particle beam. This maximizes payload possibilities, and there is the possibility of using magsail technologies for deceleration. Icarus, however, is a fusion project.

    As to aerobraking, the problem there is that a craft moving at a substantial percentage of c is in imminent danger of destruction by collision with even tiny objects as it enters its target system. Even if adequately shielded, it would still be moving far too fast for any attempted aerobraking to slow it into a stellar orbit.

  • george scaglione March 26, 2009, 9:34

    howard i have no doubt that the information you suggest above could be returned with the obvious “problem” that if the craft was 5 light years away we would expect to have to wait 5 years for the data.lol again….ideas like star trek’s sub space transmissions begin to look real good! thank you very much your friend george

  • ljk March 26, 2009, 11:07

    Would we be able to use any large planets in the target star
    system, or even the star itself, to slow Icarus down in a kind
    of “reverse” gravitational slingshot effect? Pass them by the
    worlds continuously to let their masses eventually slow down
    Icarus enough to settle into at least a solar orbit? Or will the
    probe be going just too darn fast for this to happen?

    I wonder if there is a way to turn the most potential hazard
    of the mission – being struck by debris, especially during the
    flight through the target star system – into a way to slow the
    vessel down enough to keep it in the star system?

    Maybe we should forgo trying to get Icarus to Alpha Centauri
    in one human life time and just accept that it will take longer,
    therefore we don’t need to be turning ourselves into pretzels
    trying to satisfy every criteria. Plus with human life spans
    increasing all the time, a 200-year mission might not be
    unreasonable by 2050 for its builders to still be around when
    Icarus gets there.

  • ljk March 26, 2009, 13:18

    The World’s Biggest Laser Powers Up

    Technology Review Mar. 26, 2009


    The National Ignition Facility (NIF), a laser system designed to produce nuclear fusion reactions that release more energy than used to produce them, is now up and running.

    By 2010 or 2011 the NIF’s 192 lasers will be able to deliver 1.8 megajoules of energy in a few billionths of a second to one 2-millimeter sphere filled with hydrogen…


  • Ron S March 26, 2009, 14:51

    ljk: “Would we be able to use any large planets in the target star
    system, or even the star itself, to slow Icarus down in a kind
    of “reverse” gravitational slingshot effect? Pass them by the
    worlds continuously to let their masses eventually slow down
    Icarus enough to settle into at least a solar orbit? Or will the
    probe be going just too darn fast for this to happen?”

    Anything moving at a substantial fraction of light speed is moving too fast. I did once calculate this, and while I don’t remember the details, the details are pretty much irrelevant. Grazing a planet puts it into its gravity well for a time measured in milliseconds, and even only seconds for grazing a star. Deflection is related to the gravitational acceleration of the body being approached, which is a puny fraction of the forward velocity. Add those two vectors and the result is pretty much identical to the original, forward vector.

  • Ron S March 26, 2009, 15:06

    Paul: “Laser dispersion is far, far less than radio. The beam is tightly collimated and Lesh and Cesarone believe the signal would be readily discernible.”

    It’s all about beam-width which is, roughly, an inverse-square relationship to gain. Apart from that, photons are photons. That is, halve the beam-width and get 4x gain (6 db). The advantage of a laser is simply that the wavelength makes achieving a narrow beam-width (high gain) easier (aka cheaper or more technologically feasible). And according to the there-ain’t-no-such-thing-as-a-free-lunch principle, you can’t increase gain without decreasing beam-width. Gain is our objective, with beam-width being the means to that end.

    Propagation through the ISM is also dispersive and will limit the minimum-achievable beam-width per unit distance. However this is more of an absolute dispersion rather than proportional to beam-width, therefore the ISM is biased against narrow beams such as lasers and acts to impose certain limits on achievable gains. I haven’t bothered with figuring out how to do an exact calculation, but perhaps the Lesh and Cesarone work you referenced deals with this.