≡ Menu

Collisions in the Interstellar Medium

Memories play tricks on us all, but trying to recall where I saw a particular image of a laser lightsail is driving me to distraction. The image showed a huge sail at the end of its journey, docked to some sort of space platform, and what defined it were the tears and holes in the giant, shredded structure. It presupposed long passage through an interstellar medium packed with hazards, and although I assumed I would have seen it on the cover of some science fiction magazine, I spent an hour yesterday scanning covers on Phil Stephensen-Payne’s wonderful Galactic Central site, but all to no avail.

The image must have run inside a magazine, then, but if so, I’m at a loss to identify it other than to say it would have appeared about twenty years ago. I had hoped to reproduce it this morning because our talk about starship shielding necessarily brought up the question of whether an enormous lightsail — some of these are conceived as being hundreds of kilometers in diameter — wouldn’t be impractical in denser areas of the galaxy. And that brought to mind a 1986 exchange between the British astronomer Ian Crawford (Birkbeck College, London) and Robert Forward, the physicist who did so much to awaken us to the possibilities of interstellar flight.

This morning I’m about eight miles away from the library where I can find back issues of the Journal of the British Interplanetary Society, but Gregory Matloff and Eugene Mallove wrote about this correspondence in The Starflight Handbook, which I do have right here in front of me. Forward’s position was that a laser lightsail would be so thin that dust grains would pass right through it without depositing a great deal of their kinetic energy as heat. So maybe the shredded lightsail isn’t a necessary outcome of a beamed sail mission. From The Starflight Handbook:

During a 10-ly journey at 0.2 c, only 1/500 of the area of a 0.0160 micron (160 Å or angstrom) thick light sail will be lost. However, Forward and we agree that a great deal of theoretical and experimental work on interstellar erosion must still occur before we can set off for the stars free of bad dreams.

Dust Grains Between the Stars

I’ve been thinking about Ian Crawford partially because of his recent paper on manned spaceflight and its virtues, but also because of another exchange he had about two years ago, this one with Jean Schneider (Paris Observatory), who had been examining our response to the detection of biosignatures on exoplanets, and in passing discussed how difficult it would be to get a probe to an exoplanet to investigate them. Schneider was worried about the interstellar medium too, and went to work on the possibilities assuming a spacecraft velocity of 30 percent of the speed of light. Moving at that pace, Schneider calculated that a 100-μm interstellar grain would have the same kinetic energy as a 100-ton body moving at 100 kilometers per hour.

The Schneider/Crawford exchange is up on the arXiv site (references below), and you can read about it in Interstellar Flight: The Case for a Probe as well as Interstellar Flight and Long-Term Optimism, the two articles I wrote about it back in 2010. It was Crawford’s position that interstellar dust grains could indeed present a hazard that will need to be factored into the design of the vehicle, but Crawford found several mitigating factors including speed, pointing out that 30 percent of lightspeed was a overly ambitious target, and certainly a more problematic one, owing to the scaling of kinetic energy with the square of the velocity.

Crawford finds that the situation at 0.1 c is considerably better. From the paper:

The issue of shielding an interstellar space probe from interstellar dust grains was considered in detail in the context of the Daedalus study by Martin (1978). Martin adopted beryllium as a potential shielding material, owing to its low density and relatively high specific heat capacity, although doubtless other materials could be considered. Following Martin’s (1978) analysis, but adopting an interstellar dust density of 6.2 x 10-24 kg m-3 (i.e., that determined by Landgraf et al., 2000), we find that erosion by interstellar dust at a velocity of 0.1 c would be expected to erode of the order of 5 kg m– 2 of shielding material over a 6-light-year flight.

We clearly need shielding, then, and that adds to the mass of the interstellar probe, but Crawford does not find the problem insurmountable. Of course, what we have yet to learn is the true size distribution of dust particles in the nearby interstellar medium, which is one reason we need a mission like Innovative Interstellar Explorer, to make such measurements in situ. Crawford works out the spatial density of 100-μm grains at about 4 x 10-17 m-3 based on work by Markus Landgraf (Johnson Space Center) and colleagues in 2000. Here we find just how much work lies ahead:

…over the 6 light-year (5.7 x 1016 m) flight considered by Martin (1978), we might expect of the order of two impacts per square meter with such large particles, and the injunction by Schneider et al. (2010) may, after all, appear pertinent. On the other hand, it is far from clear that it is valid to extrapolate the distribution to such large masses, not least because of the difficulty of reconciling the presence of such large solid particles in the LIC with constraints imposed by the cosmic abundances of the elements (as also noted by Landgraf et al., 2000, and Draine, 2009). Clearly, more work needs to be done to better determine the upper limit to size distribution of interstellar dust grains in the local interstellar medium.

More work indeed, and Schneider, in a response to Crawford’s own response to his earlier paper, notes that the matter comes down to what we are willing to live with in terms of probabilities:

The question is what probability of collision is acceptable. If a collision is lethal, this probability must be extremely close to zero for a several hundred billion € mission.

Searching for Solutions

We’re not yet able to make the definitive call on just how many large interstellar grains our probe may run into in the local interstellar medium, but Crawford thinks the problem can be addressed through the detection of incoming large grains and the use of either laser or electromagnetic means to destroy or deflect them before they impact the spacecraft. Again I turn back to Alan Bond’s idea of a dust cloud ejected from the vehicle and preceding it along its course. Remember that Bond was working with the Daedalus concept of an initial, multi-year period of acceleration followed by decades of coasting to reach Barnard’s Star. A dust cloud like this could destroy larger interstellar grains before they ever reached the main vehicle. Adds Crawford:

This concept was developed for Daedalus in the context of protecting the vehicle in the denser interplanetary environment of a target star system, but it would work just as well for the interstellar phase of the mission should further research identify the need for such protection.

A mature space exploration infrastructure here in our own Solar System is probably the prerequisite for the kind of interstellar probe Crawford is talking about, and he notes the value of building that infrastructure not only in terms of creating the technologies we’ll need to get to the stars, but also in terms of making possible the search for life not only on Mars but further out in the system. How long it takes us to build this framework plays directly to Schneider’s point that:

It is presumptuous to predict exactly what will happen after one century and into the future, but it is more than likely that development of the capacity to observe the morphology of meter-sized organisms on exoplanets will take several centuries, at least in the framework of present and forseeable physical concepts. Another optimistic possibility would be that, in a nearer future, we will detect pictures of extraterrestrials with a good resolution in SETI signals. The debate must still go on.

The initial paper by Jean Schneider is “The Far Future of Exoplanet Direct Characterization,” Astrobiology Vol. 10, Issue 1 (22 March, 2010), available as a preprint. Ian Crawford responded in “A Comment on ‘The Far Future of Exoplanet Direct Characterization’ — the Case for Interstellar Space Probes,” Astrobiology Vol. 10, Number 8 (2010), pp. 853-856 (preprint). Schneider’s follow-up response to Crawford is “Reply to « A Comment on ”The Far Future of Exoplanet Direct Characterization” – the Case for Interstellar Space Probes » by I. Crawford,” Astrobiology Volume 10, No. 8 (2010). I don’t have the page numbers on the latter but the preprint is available.


Comments on this entry are closed.

  • Eniac April 13, 2012, 10:47

    There is 100 times more gas in the ISM than dust, and the gas will likely deposit more of its energy than dust, especially when passing through a thin sail. At near relativistic energy, I would think, each atom of gas would scatter many atoms of ship, and thus be very corrosive. Why are we focusing on dust grains, then? Am I mistaken?

  • A. A. Jackson April 13, 2012, 11:01

    Just to note the ESA Ulysses has spent , now, nearly 20 years making dust measurements in the Solar System (it makes many other measurements too).
    I think the interstellar dust distribution in the vicinity of the Sun is pretty well known.

  • Christopher L. Bennett April 13, 2012, 12:08

    Projecting a dust cloud ahead of the ship as a “navigational deflector” is an interesting idea, but it seems wasteful, because I’d think the dust cloud would spread out and dissipate and would need to be replenished. Ideally you don’t want your ship to carry any more mass than it needs to, so if you have to carry a big tank of replacement dust, that would be bad.

    I’m put in mind of something I’ve read about before on this blog, a type of magnetic (magnetohydrodynamic?) force field that can “lock” objects in place relative to each other. Could this be used to hold a cloud of small particles in place at a fixed distance in front of a ship, far enough out to absorb the impact of any oncoming particles? Or would the necessary distance be too great? Failing that, could lasers be used as “optical tweezers” to hold a cloud of, say, tiny glass beads in place at a fixed distance in front of the ship? I’m not sure that could work, since they’d need to pull as well as push.

  • Paul Gilster April 13, 2012, 13:02

    Al Jackson writes:

    Just to note the ESA Ulysses has spent , now, nearly 20 years making dust measurements in the Solar System (it makes many other measurements too).
    I think the interstellar dust distribution in the vicinity of the Sun is pretty well known.

    True enough, but we don’t know nearly enough about what happens once we move outside the heliosphere, which is why I’m so much in favor of Innovative Interstellar Explorer.

    Eniac writes:

    At near relativistic energy, I would think, each atom of gas would scatter many atoms of ship, and thus be very corrosive. Why are we focusing on dust grains, then? Am I mistaken?

    I know one or more members of the Icarus team who are thinking about starship shielding will have read this post by now, and I’m hoping we can get a comment from them. I know Matloff has considered the flux of radiation induced by moving through interstellar hydrogen in relation to the shielding a manned vehicle would need, but I don’t know about the absorption of energy from gas into the sail.

  • Eniac April 13, 2012, 13:54

    I recall once having calculated the energy deposited by ISM gas per area on a shield, and at 0.3 c, I believe, it was up to where the shield would have to be quite hot to radiate it all back into space. I suspect erosion is much worse than energy deposited, but I don’t really know. There seems very limited work on erosion by relativistic particles. An occasional article about the longevity of accelerator targets, but I found nothing that was readily applicable here.

    Intuition tells me that a particle going at such high velocity will knock hundreds or thousands of ship atoms out of place with enough force for them to be lost, but I am too rusty to try to calculate that. On the other hand, I calculate that the ISM has about a few micrometers worth of material (were it condensed) between here and Alpha Centauri, so between the two I would expect a shield erosion of the order of millimeters for that trip. Good enough for a low cross-section macroscopic probe, but no good for a sail or a “nano-probe”.

  • Eniac April 13, 2012, 14:11

    I suppose Rutherford scattering could be applied, and there would be two limit cases to be considered: thin shield (much thinner than penetration depth, like a sail) and thick shield (thicker than penetration depth, a full shield). You’d calculate number of scattering events per incoming gas atom, how many would impart sufficient energy to knock shield atoms out of place, and how many of those atoms would emerge through the surface. A fairly sophisticated calculation, but not beyond a thesis among the lines of the recent one on FOCAL (hint, hint).

    Plus, you could borrow the LHC for some experiments, after they confirm the Higgs boson :-)

  • Michael April 13, 2012, 15:56

    A possiblity for the front shielding is instead of using a hard material use a softer denser one that is in multiple layers or a liquid that could be moved to densify another region at high speed pre-empting a high mass impact object, if the sputtered material could be recollected it could be reused to repair the shield. Ultimately a combination of techniques maybe be more suitable, all techniques must be mindful of mass requirements though.

  • Alex Tolley April 13, 2012, 17:39

    Isn’t the earth a model for relativistic impacts of gas and dust from high energy events? Presumably a column of gas 100km high under 1 g compression is a good shield. So can we replicate that (as suggested by other commentators) in some way to provide adequate shielding for the vessel? The mass requirement seems high (1 kg/cm -2) but maybe something less extreme would work too with adequate protection.

  • Jim Early April 13, 2012, 17:53

    Eniac is correct; the gas is approximately 100 times the mass of dust in the local ISM. That ratio is lower in the nearby “local bubble” where much of the gas has been removed, presumably by a super-nova. If the gas is ionized, then at 0.1c a simple shield charged at 15MeV will deflect any ions. The more problematic neutral gas atoms can be ionized first by having them pass through a thin foil placed in front of the ion deflector.

  • M. Lades April 13, 2012, 20:50

    Looking for classic Sci-Fi: Just like you I am remembering a book from the late 60s at best early 70s, that described a mission into the interstellar medium, where the two ships are turning around basically because the interstellar environment is described as too hostile by aliens encountered. They book used various drive stages including photon for the interstellar environment. It described Sanger like shift of colors ahead and behind the ships. A person like a “Commodore Hagen” may have played a role. Could have been English or German. Any pointers appreciated.

  • Eniac April 14, 2012, 5:12

    Jim, I understand that at least in our area, and most everywhere else, the interstellar gas is not ionized. So, a foil ionizer plus electrostatic or magnetic deflectors may be the way to go. I suppose that the foil will need to be continuously replaced, because of erosion. Or not?

  • Michael April 14, 2012, 5:46

    @Jim Early

    The problem with a charged shield is that it will repel same charges but attract the opposite ones, having said that I would rather be hit by an electron than a nucleus though. Neutral gas can be ionised by lasers or high frequency oscillating fields or evening moving through a magnetic/electric field of sufficient strength.

    On another note it is a pity that there was not more interaction between the Icarus team and the forum, as can be seen here there are plenty of ideas and interest.
    Further could we not have a tick up icon on a remark for good ideas that could be looked at the end of a set period or we could just have a vote on the best idea or concept of the year, it seems good ideas just disappear into the ether.

  • Paul Gilster April 14, 2012, 8:41

    Michael writes:

    Further could we not have a tick up icon on a remark for good ideas that could be looked at the end of a set period or we could just have a vote on the best idea or concept of the year, it seems good ideas just disappear into the ether.

    I’ll be glad to explore any ways that can keep good ideas available. Keep in mind, too, that everything written on Centauri Dreams since day one is available in the archives, so keeping it out there as a research repository means that these comments don’t disappear, and are available to researchers.

  • Abelard Lindsey April 14, 2012, 12:09

    So, 10% of light speed appears to be the limit, eh?

  • Peter Popov April 14, 2012, 15:25

    your idea of a liquid, or thin, layered structure seems like the best idea for shielding against dust particles. I guess you could get a honeycomb frame for support and fill the voids with a thin layer of solidifying liquid. So you have a segmented foil. If a dust particle hits such a foil “plate”, both are destroyed. The frame prevents the impact from affecting the rest of the shield. After such an event, a thin layer of liquid is spread over the affected area, which then solidifies. The hole shields is kept some good distance away from the main probe.

    I cannot quite imagine a dust cloud being maintained at sufficient density without much material loss. I also cannot imagine an electronic system detecting small object and destroying them actively (which will be relevant at the target star system). If you through in enough power into a radar, you probably could, but then do you have enough power for the entire voyage.

    Clearly though, the shielding problem is easily tractable with present day technology. So , I am amused by the amount of energy that goes into discussing easily solvable technical challenges which are uniquely relevant to a starship. Devising a fusion propulsion system is the real problem. A fusion drive has many other uses (interplanetary travel, power generation, etc). So, directing one’s energy at this problem has much more impact than figuring out in great detail the shielding of a hypothetical probe, that may never be built.

  • Ole Burde April 14, 2012, 16:05

    In a rescent post we read about plasmawawe propulsion , where no “fuel” is used , but where the density of the surounding plasma could be critical for efficiency . If a spacecraft can protect itself from dust and gas by turning them into plasma , and deflecting the produced plasma just enough to “get out of the way” , this might be exactly what is neded to make plasmawawe propulsion possible .

  • Michael April 14, 2012, 17:07

    @Peter Popov

    This is the idea I had in mind for dealing primarly with dust, perhaps the laser that was used for dispersing gases ahead could be used to illuminate dust particles ahead at which point the power of the laser would increased significantly either destroying it and giving time for a liquid such as the link displays to increase in depth before impact -admitantly it would have to be quick, http://t3.gstatic.com/images?q=tbn:ANd9GcQ7eJmfJS49bwgSvs2QD3eGLCGhAV
    The liquid would then absorb the impact and heat up radiating the energy and reforming the shield, the dust absorbed now becomes part of the shield adding to the protection.

  • Michael April 14, 2012, 17:28
  • bill April 15, 2012, 14:47


    ” I recall once having calculated the energy deposited by ISM gas per area on a shield, and at 0.3 c, I believe, it was up to where the shield would have to be quite hot to radiate it all back into space.”

    Would you by chance be able to reproduce the calculations here ?

  • bill April 15, 2012, 14:56

    a interesting Adobe pdf on accelerator target cooling; might apply here …


  • Rob Henry April 15, 2012, 19:02

    Bill, these are not esoteric calculations, requiring some deep understanding of the Napier Stoke equations. These molecules are moving so fast that the energy required to remove them from the path of the shield in real time could have little difference from their total kinetic energy. This is just ½ mv^2. The mass of the ISM to be thrown aside would vary as it’s density times the speed of the craft. So (forgetting relativity since gamma is only 1.05 here) the power to be dissipated per area is just the ISM density times half the cube of velocity.

    At the commonly given 1 molecule per cubic cm this is 80 kW/m^2. It should be radiating at over 1000K. Disturbingly, in a “dense” molecular cloud it would have to radiate more like 80 GW/m^2.

    I think that, while Eniac wants to show how this is a problem in itself, it is trivial in comparison of how you prevent a particle (usually H) impacting at so many thousands of times more energy than any molecular bond, can be prevented form knocking several molecules of that shield (built of much heavier elements). I also have a horrible feeling that at these speeds you would be doing well if you lose just 100kg of shield per kg of neutral gas displaced.

  • Eniac April 15, 2012, 19:15


    Would you by chance be able to reproduce the calculations here ?

    What there was is reproduced here:
    Except I now realize that the part about radiation temperature has been left as an ‘exercise for the reader’. Use Stephan/Boltzmann… Even without, Megawatts per square meter says red hot to me.

  • Rob Henry April 15, 2012, 21:59

    In those earlier calculations by Eniac I note that the velocity is about three times as great when the effects of relativity are taken into account, so my figures have to be multiplied by about 20-30 for comparisons. Thus our results are effectively the same, they just look very different.

  • Brett Bellmore April 15, 2012, 22:40

    “it is trivial in comparison of how you prevent a particle (usually H) impacting at so many thousands of times more energy than any molecular bond, can be prevented form knocking several molecules of that shield (built of much heavier elements).”

    I’d presume that, if we’re talking molecular particles, that they’d penetrate beyond the surface of the shield before depositing most of their energy, and then the deposited energy would carry with it momentum which would cause the recoiling shield atoms to be driven further into the shield. So your loss would mostly consist of evaporation off the surface of the shield. Which could be pretty serious at red heat over decades…

  • jkittle April 16, 2012, 8:20

    For O.B.
    note that the concentration of H3+ is significant in the interstellar medium ( see my late post in the previous blog entry) . underdatd though that we would need REALLY big magnetic fields these take a lot of energy to construct

  • Ole Burde April 16, 2012, 16:55

    A veldesigned boat sailng through water use very little energy to moove the water around its shape , because it gets the the energy back from the water , when the water flows back to its original position . A starship deflecting the ISM should be capable (theoreticly) to do the same , even if the forces are of a completely different kind . It would be a lot easyer to imagine for a small compact structure .

  • Bounty April 16, 2012, 19:51

    What is the difference between putting a (open ended?) tank of water at the front of the ship and the dust cloud idea? I imagine as you increased in speed, the water would get hit, heat up and expand forward (possibly still in a large cylinder or contained in other ways) as a gas acting as your shield. Perhaps it’s ionized and contained that way. You might need a huge radiator, but you’re probalby already doing that for the engine that gets you up to that speed? Also you could vent the extremely hot gas (hopefully in a useful way or use some of that energy) If not, the hottest gast would just expand away once it got past containment. The ship could be be straw or pencil shaped, limiting it’s cross section.

  • Bounty April 16, 2012, 19:57

    Also, cruise, acceleration and deceleration phases might look very different as far as problems with shielding go.