Talk:Interstellar travel/Archive for 2010

Difficulties

I have just reverted an edit by User:Paulkint on the difficulties. The edit was unsourced and formally original research, though probably true. However its appropriateness is controversial, as we have discussed here on several previous occasions. I think this material has been pushed to show that interstellar travel is "impossible" (which is not established), not merely "difficult" (which is not controversial). Of course the distinction is a value judgment, about which people may disagree. But the opinions of us editors, in the absence of consensus, is not sufficient. Thus I request reliable external sources for any such alternations. Thanks. Wwheaton (talk) 02:17, 19 February 2010 (UTC)

Interstellar ramjets

The answer to a question lately put into the article ("Is there drag in space") by User:207.189.243.126, and later reverted, is that "it depends". In general the drag in space is very small, but if you built a ramjet (see Bussard ramjet) big enough to produce significant thrust, with a huge scoop, then it would almost certainly have significant drag as well, and I believe that it would be extremely difficult to come up with a configuration in which the thrust would be greater than the drag. As far as I know, no design concepts that seem remotely feasible (without technical magic) have been proposed. Wwheaton (talk) 06:33, 20 February 2010 (UTC)

Seconded. This was one of the big objections to Bussard's original ramjet proposal. It's still been used widely in fiction, of course. For non-ramjets, though, drag is generally negligible (the ship greatly outmasses the cylinder of interstellar medium it travels through). A back of the envelope calculation based on interstellar medium gives about 20 mg of ISM per square metre of area per light-year traversed on average, and about a tenth that in the Sun's neighbourhood (we're in a low-density bubble region). --Christopher Thomas (talk) 09:01, 20 February 2010 (UTC)

One could imagine the gas being adiabatically (& reversibly) compressed, burned, and then losslessly re-expanded, which would not have to create drag in principle; but in practice a fusion plasma of the temperature and density needed to burn ordinary hydrogen (besides requiring impossibly high temperatures and pressures) would radiate X rays and gamma rays like the core of a supernova, and thus could not be adiabatic or lossless without some kind of perfectly reflecting (for X rays and gamma rays) containment vessel. Wwheaton (talk) 16:53, 20 February 2010 (UTC)

A variant that works (albeit poorly) is to consider a drive that halts (brings to craft speed) all hydrogen striking it, sorts out the deuterium, burns that, and vents the exhaust. If you throw away the light hydrogen after bringing it to craft speed, and vent the fusion plasma as rocket exhaust, you get a maximum speed (when thrust equals drag) too low to be useful (about 6 km/sec). If you instead mix the light hydrogen with the post-reaction fusion plasma, then vent it, you get a possibly-useful maximum speed of about 370 km/sec. I'm assuming that only energy held by charged reaction products is recovered (about 2.5 MeV/nucleon for D+D fusion). Radiative losses are avoided for this scheme the same way they are for a magnetic confinement fusion reactor: confinement time is short enough that radiative transport doesn't carry away much of the energy.

In practice, 1) this is an upper bound (assuming no wasted energy), and 2) you can build a 400 km/sec fission-electric drive much more easily than a 400 km/sec fusion drive, so even if someone invents a practical D+D fusion reactor, building a ramjet with it is unlikely to be attractive. It's just fun to work the numbers! --Christopher Thomas (talk) 17:35, 20 February 2010 (UTC)

NASA research section

I just fixed some formatting problems and a broken link in this section, but after doing that, I deleted some interpretive commentary that seemed indefensible. The claim that a "necessary" condition for I.S. travel is "A method of propulsion able to reach the maximum speed which it is possible to attain" appears absurd, since Special Relativity tells us c cannot be reached, and all Lorentz frames with v < c are physically equivalent. There are of course practical and economic constraints, but these are never absolute hard limits, but only gray regions where the difficulties become more and more overwhelming. The whole NASA Glenn BPP Project pages are really fairly good, or at least a lot of good material and references are in them, but I did not find the two "criteria" that were listed in a cursory look-through, so I think my deletions were of unwarranted extrapolation of the refs. (Well at one place it does say we "need" FTL speeds, yet gives little or no reason for believing that these are not fundamentally impossible; I can't take that very seriously either.) Probably this whole section needs to be cherry-picked for good material, and then simply reduced to a reference or two (or three...). Wwheaton (talk) 06:11, 2 April 2010 (UTC)

At some point, if I have the time, I'll look through the NASA report itself (rather than press releases about it). I'm pretty sure the hyperbole in the additions about it were an artifact of the assumptions made by the BPP researchers. If they were assuming that the only trips of interest were ones that would be performed in 20 years or less, then they would correctly conclude that all conventional means of propulsion are inadequate. My impression from reading about it a while back was that it wasn't a serious attempt to look at drives that could be built using extensions of existing technology, but an effort to come up with exotic ideas for propulsion (of which the diametric drive was the most interesting; the bias and pitch drives all made use of serious handwaving). Given that goal, it's natural that they'd choose their starting assumptions so as to result in a need for such drives.

That said, these impressions are based on second- and third-hand reports. Hence, reading the document itself one of these years. --Christopher Thomas (talk) 06:36, 2 April 2010 (UTC)

Black hole Hawking radiation

Doesn't "Black hole Hawking radiation" belong to "further speculative methods"? Or is it too speculative to be in the article at all? Offliner (talk) 23:29, 25 November 2009 (UTC)

Yes. I've moved it. Regarding the concept itself, the idea of using a parabolic reflector is also pretty iffy, as a decaying black hole will emit particles and antiparticles in addition to photons, and for small holes most of the photons will be very high-energy (hard to reflect). I'd go for a magnetic nozzle and a thick shield rather than trying to build a reflector. Acquiring the black hole is left as an exercise. --Christopher Thomas (talk) 23:41, 25 November 2009 (UTC)

Of course it has the ravishing beauty that if Hawking is right, then a BH converts mass into radiation with 100% efficiency, and if its mass were regulated properly (by carefully feeding it more mass, presumably) it could radiate in a region (say in the submillimeter, T ~ 0.1 K?) where a superconducting reflector could be practically perfect. I think the Q of S/C cavities can be upwards of 10¹⁰, if I recall. I don't know of any other power source nearly as good. (Matter/antimatter annihilation is always mentioned, but is quite distasteful due to gamma-ray and neutrino losses.) We would probably have to find an appropriate mass BH, and then feed or starve it to "tune" its temperature. The temperature would have to be quite low, cryogenic, and the mass large; less than a solar mass, maybe something like a terrestrial planet mass. And at such a low T, the black body luminosity of a <1 cm radius BH would seem to be uselessly low. (Actually, I guess it must have T > 2.73 K to prevent it from gaining energy, from the CMB, faster than it radiates it.) But not fundamentally impossible?? Wwheaton (talk) 07:00, 2 April 2010 (UTC)

A superconductor is actually a less-than-ideal conductor for alternating current, and so a less-than-perfect reflector, even at radio/microwave frequencies. Still quite good, but then so is aluminum or copper. If I were building a photon drive based around a black hole, I'd probably use very thin gold foil and have it emit in the mid-infrared.

The difficulty, though, is that it has to be emitting brightly enough that you feed much more than the mass of the hole in reaction mass through it over the course of the burn period. If we assume we want one black hole's worth of mass to vanish in a year, we can get a handwavy estimate (a bit high) of the black hole's maximum mass by looking at the mass of a hole that evaporates in a year. From the Hawking radiation article, evaporation time is about 8.4e-17 times the cube of the mass, so for 3.2e+7 seconds you get a maximum mass of about 7.2e+7 kg (actually somewhat less than this). The temperature of a black hole is about 1.23e+23 / M, and the peak thermal emission wavelength is about 2.4e-26 M, so the heaviest (coolest) feasible hole would have an effective temperature of about 1.7e+15 K, and a peak thermal emission wavelength of about 1.7e-18 m, if I'm doing the math correctly.

So it looks like we're stuck with a hole that produces particle/antiparticle pairs for everything up to about the 100 GeV range, with this exhaust deflected the same way we'd deflect a matter/antimatter plume. Considerably easier to get the fuel than with an antimatter drive, though stuffing it into the hole might be problematic (it's small, so most things will tend to do a hyperbolic orbit around it instead of falling in). I'm also not sure how much of this comes out as matter and how much of this comes out as gamma rays; gamma ray heating would be another serious limitation to the drive.

Interesting to think about, though. --Christopher Thomas (talk) 07:58, 2 April 2010 (UTC)

The Q of superconducting gold microwave cavities can be so high (I believe--I did this calculation years ago--gotta go check about superconducting RF accelerators) that a layer of Au just a few skin thicknesses deep could be accelerated at >0.1g by radiation pressure. If that is actually correct we might have our reaction "motor", and "only" need a suitably efficient, low-mass power source. I agree that seems to be a killer for the BH, in the sense that the power-to-mass ratio at microwave emission frequencies looks hopelessly low, and I think we don't know any way to cope with the catastrophic energy losses and inefficiency at higher frequencies. Yes, fun to think about! Wwheaton (talk) 16:50, 2 April 2010 (UTC)

Constant acceleration drive

This section notes an important property of special relativity that has been known for decades, and which probably merits a paragraph of mention in this article, though it is not really a "method" or a "drive". The point is that because of the Lorentz time contraction factor, the distance attained (and average speed, measured on-board), increase exponentially with ship time, for uniform acceleration. The wiki article on rapidity describes some of the math, but "rapidity" is essentially the speed (measured in speed-of-light units, relative to c) a naive Newtonian spacefarer would believe he had reached if he never looked outside his ship and just relied on "dead-reckoning", or an inertial guidance unit. The really important thing for this article is that special relativity actually enables extremely fast FTL, eg, billions of light years in a human lifetime—measured on the ship. The problems, or "gotchas", are two: first, if you travel billions of light years, then billions of years will have passed on Earth, and you can never go back to the era from which you departed; and second, nobody knows in practice how to build a ship that can go that fast, or attain a rapidity of even 0.1 (very nearly a speed of 0.1c, at such low speeds), let alone a rapidity of 1 or 35. In any case, achievable rapidities are limited by the same old rocket equation that bugs space travelers today, so that 100 km/s is still beyond us. Note also that, as rapidity is simply additive (unlike speed in relativity), there is no magic associated with the "constant acceleration" strategy, although it may be optimal for some circumstances. In general, you will go farthest by coming up to top speed as fast as possible (ie, as fast as your ship and g-tolerance allow), then coasting, when braking to a stop as fast as possible.

I think the current section is far enough off the mark that it needs to be removed, but probably the above facts deserve some place in our article. For now I will just comment it out, and wait for comment from others. I first learned about it around 1955 in an old book by Arthur C. Clarke (probably The Exploration of Space), and he was quoting an earlier worker in the field, so it is not at all new, though I think it is marvelous for the crack in the wall of impossibility it reveals. Wwheaton (talk) 00:06, 5 April 2010 (UTC)

I think that the concept does deserve to be mentioned, but that the subsection about it should be kept short. There really isn't much to say about the concept itself: Just that by travelling arbitrarily close to the speed of light, you can make the travel time as short as you like. We could state and reference that fact (probably from a GR text that discusses hypothetical space travel, as it's a common example), and state that it's been widely used in fiction (citing a couple of the most noteworthy examples; The Forever War being the best that comes to mind, with others existing but the ones that I've read (like A World Out of Time) not being terribly noteworthy IMO). This would be one paragraph, at most.

What we'd probably have trouble referencing is statements about the very large energy costs required (many times the rest mass of the ship, by a factor roughly proportional to the time compression at large compression).

As this exceeds the rest mass of the ship, it couldn't be provided by fuel carried by the ship. If provided by a gun-like mechanism, the mechanism required becomes extremely long at survivable accelerations (about N light-years for a time compression factor of N at one gravity; K times less at K gravities). If provided by a beamed propulsion system, the efficiency drops quite a lot due to redshift (to look at it another way, you're almost outpacing the drive beam, so even from the viewpoint of an observer at rest with respect to the drive station fewer photons/ions/etc strike your sail per unit time than when you're at low speed). This magnifies an already-vast amount of energy further. So, I seriously doubt that using relativistic time compression to shorten subjective travel time will ever be an attractive proposition, as it'd be far easier to dump the resources into establishing reliable suspended animation or mind-upload technology. At most, I could see similar systems being attractive for very small unmanned probes sent with very high acceleration, but these wouldn't have to get to high time compression (just close enough to C that further money spent on infrastructure doesn't get enough of a decrease in non-dilated travel time to be justified).

Long story short, if the subject gets more than a paragraph, it's overkill, and there's only one aspect I think we'll have trouble finding references for. --Christopher Thomas (talk) 04:10, 5 April 2010 (UTC)

Well, it interests me that it really does allow one to go farther and faster with relativity than without it. In a Newtonian universe, if you accelerated at 1 g for 30 years of ship time (meaning the non-relativistic rocket equation would make you engineer the rocket for that, 30 years) then you would reach a maximum speed of ~30c, and go about 15 light years. But in the real universe, with special relativity, you go much much farther, beyond the most distant galaxies I think, though I'd have to check the math to get the exact distance. If I recall the distance reached in ship time τ involves a factor of [cosh(aτ/c)-1], where a is acceleration, c is the speed of light, and τ proper-time, meaning the time by normal clocks aboard the ship. This factor is around 1.e13 if τ = 30y and a = 1 g, which means your perceived speed far exceeds c, partly because your clocks are running slow, and partly because the Lorentz contraction shortens the distances you have to cover.

BTW, I found the earlier discussion in the 2009 archive; the second thread, "Sub-light travel" has a derivation by someone named Steve". that I have not checked, but which looks roughly like something I calculated several years back. This was apparently done before the signature/date stamping of talk page entries was instituted. And then there is some more comment (some by me) in Archive 2008, written later but archived earlier, in section 5 called "Super light speed" travel, written between March and August 2008. I thought at the time that "Steve"'s calculations were basically correct, but I'll have to go back to the books to be certain in detail.

Anyhow, the bottom line for me is that relativity actually helps, and allows us to beat the Newtonian case in principle, and by very large factors. BUT—it does not kick in until we get up near relativistic velocities, which we cannot do with known means of propulsion. Wwheaton (talk) 07:08, 5 April 2010 (UTC)

My point is that this only "helps" in terms of reducing the amount of subjective travel time, and that there are vastly less expensive ways of accomplishing that goal. In terms of absolute energy or effort or infrastructure required, it ends up making things far worse (compared to, say, travel at 0.7C).

To put it another way, why would I spend a million times your rest mass to boost your ship to 0.999999999999C for a trip to Andromeda, when I could spend 3 times your rest mass to get you to 0.9 C, and freeze you for the trip? Sure, the maintenance systems on the ship will experience different amounts of time (3 years vs. 1 million years), but you as a passenger won't care, and from my point of view you still take about 2.5-3 million years to make the trip either way. Vastly less effort, with no downside.--Christopher Thomas (talk) 18:38, 5 April 2010 (UTC)

Well,I guess I'm just saying that it removes the speed of light as a "barrier" if you want to go somewhere far away and don't plan to return, and actually allows long trips with less rocket performance than Newtonian physics. I suppose I think it would be worthwhile to dispel the misconception that we cannot travel to the stars because we cannot go faster than light.

My current interest is just in trying to figure how to get human colonists to nearby stars in the "near term", by which I mean (in this context) less than 1000 years, say, without magical intervention, at ~0.1c or so. The aim being to secure the long-term survival of Earth life, or its descendants, given the major hazards that we all know too well on millennial time scales. If course if what someone really wants is to zip off to the Virgo Cluster to buy better hamburger, all these discussion are irrelevant, I don't think we can help. :) Wwheaton (talk) 23:32, 5 April 2010 (UTC)

My point is that this doesn't give you better performance than Newtonian mechanics. It only seems that way to someone riding on the ship (and even then only because they're getting their fuel for free). To outside observers, who are the ones providing the motive power to the ship, it looks much worse. Do you see where I'm coming from with this?

Regarding travel to other stars at 0.1c, I think there are at least three viable approaches, which could be done with near-present technology but would be very expensive. This was discussed at length in archived threads, so I won't restate it here. Long story short, I doubt we'll find it attractive to sink that kind of money into such a project any time soon. For survival of Earth-derived life, there are a large nubmer of places we can colonize in-system for far less effort (though they'd be more vulnerable to deliberate destruction than interstellar colonies). --Christopher Thomas (talk) 01:26, 6 April 2010 (UTC)

Gentlemen, I think you're missing the point which I was aiming for when I included this section: which is to inform the reader that the constant acceleration drive concept exists, and what its characteristics are. As I note in the first sentence of this piece, this part is not about feasibility. This section is important because it is real (as versus, say, warp drive) and because it is not well understood. In particular, the concept that ship time and planetary time will flow differently is missed by a lot of readers, which means they miss understanding the benefits to the ship travelers. By the way, the way around the providing fuel problem is to be like a sailing ship -- grab energy from the surrounding environment. I'm not saying that's easy to do, but that's the breakthrough that would make this feasible. So, let's get this back in, in some form. And, as a bit of self-promotion, I have written a whole series of SF stories based on constant acceleration technology -- The Honeycomb Comet -- as well as my article which I have cited. So, I'd like to be considered as a reference on this. Cyreenik (talk) 03:49, 6 April 2010 (UTC)

Regarding references, while I respect the fact the effort that went into these novels, I'd be reluctant to include them as references for the same reason I wouldn't include Niven's story as a reference: Neither is likely as notable as Forever War, and we should seek references in order of descending impact on the field. Finding one or two more books of similar impact to Forever War would be sufficient for establishing that the concept is notable among science fiction writers.

Regarding content of the section in the article, my view about what should be in it is already described above (the first two paragraphs of my first response in this thread). This seems to be compatible with your own desires; the problem, per my original post, is that we need to find non-fiction references before reinserting it (arguably quite a lot of other material in the article falls into this category too, but one step at a time). Everything after that reply was myself and User:Wwheaton speculating (arguably off-topic, so I agree that it should wind down in the near future).

Regarding this being "real", it's a lot less real than, say, Forward's scheme for beamed propulsion. Drawing energy from your environment doesn't work, because you're moving (very quickly!) with respect to everything that emitted light (stars, cosmic microwave background, etc). This means light striking your front has more momentum (due to blue-shift), and light striking from behind you has less (due to red-shift). Net effect: Drag, and lots of it. If you try building a big sail/solar-cell/etc to tap power from incident light, you instead end up slowing down. The exception is to use a driving beam, which is already discussed above. Regarding drawing power from something other than light, I'd call that at least as magical as using negative mass, so I'm not sure I'd accept an argument that it's more "real" than space-warping methods of rapid travel. Did you have something specific in mind for this? --Christopher Thomas (talk) 05:13, 6 April 2010 (UTC)

If you are uneasy about extensive explaining of constant acceleration in this article, perhaps we should set up a separate article and point to that? It is a concept that should be discussed for two reasons: first, it does bring out the effects of time dilation so vividly, second, it's unlikely we are going to see interstellar space commerce -- meaning thousands of ships moving routinely between star systems -- without it. The other sub-lightspeed methods such as Forward's are too slow. As for how it would work, you are being negative. You are being like saying you can't use wind power to move a ship upwind. If you said that you would have been absolutely right for thousands of years, then surprise! I'm not going to tell you how it's going to happen, but I will say that until it does happen, we won't have interstellar space commerce. This is why it's important to talk about. Cyreenik (talk) 03:36, 7 April 2010 (UTC)

What's to "extensively" explain? It can be summed up in one paragraph, given above. Heck, two sentences would probably do it: "Travelling close to the speed of light can reduce the subjective time experienced by travellers by arbitrary amounts, but energy requirements increase in direct proportion to time compression. This concept has been widely used in fiction, most significantly in (A), (B), and (C).". What are you proposing to add beyond this?

Regarding it making travel viable - it doesn't actually change anything at all. Inert cargo doesn't care what the time compression is, and humans in cryosleep or what-have-you don't care what the time compression is. The departure and arrival times are virtually the same whether I'm travelling at 0.9c or 0.9999+c. Where's the benefit gained from spending a vast amount of energy to change the amount of time the clock on your cryogenics pod experiences? If you're going to make a statement like that in the article, you're going to have to back it up.

Regarding being negative, I repeat: how exactly are you planning to "draw power" from your environment? All of the other proposals in the article have at least some basis in known physics, so yours would have to as well. --Christopher Thomas (talk) 04:43, 7 April 2010 (UTC)

Chris, as the ad goes, "Have it your way." I've had my say. Put up what you think is appropriate. If you are interested in what is to extensively explain, check out my articles at and my SF story You ask about how I'm going to draw power. I will ask you the same thing about your favorite, "How are you going to revive stiffs that have been frozen solid?" At this stage doing that is not any more feasible than drawing power from the environment. To further answer how would I draw power, note that cosmic rays manage to travel at near-light speed without getting significantly slowed down by the effects you have mentioned. But... I'm really not here to have a bull session in the discussion page. So, since this seems to be your page, you decide how much to mention. If at some point you decide you want help discussing this, let me know. Cyreenik (talk) 04:29, 8 April 2010 (UTC)

Nuclear rocket concepts revised, mainly to add "Discovery II" study

I just discovered a fairly old (2001) systems study for a fusion-powered interplanetary transportation system, remarkably similar to the Discovery vehicle of "2001" fame here. Probably it is old hat to many of you. But it seems to me to be quite comprehensive, done by (I presume) competent folks at NASA Glenn Research Center, and fills a glaring hole in our current article regarding the current "state-of-the-art", such as it is, in the high-speed nuclear vehicle design area (its characteristic velocity is >300 km/s, acceleration >1.7×10^-3g). I would be very interested to hear what other editors think of it. I think that it is not especially well-positioned for Solar System travel, where nuclear-fission/ion-electric systems seem perfectly adequate given that fusion has not yet been demonstrated even here on Earth, and the Dawn ion-drive mission is already cruising among the asteroids. Nevertheless, as fusion looks to be important as one likely piece in the "Great Interstellar Jigsaw Puzzle", I thought it deserved mention here. I have also done some re-organization and tidying up in the broader "Proposed methods of interstellar travel" section, which I hope will be inoffensive (as usual, good references are available, but often TBS). Wwheaton (talk) 01:10, 2 August 2010 (UTC)

The most famous design studies along these lines were Project Daedalus (by the British Interplanetary Society) and Project Longshot (by NASA). Having one that's more recent is certainly nice, of course, even if it's an interplanetary rather than interstellar design. 300 kps gets you one light-year per thousand years, which is in-line with my back-of-the-envelope estimates for fission-electric. The ship as-designed can do much better by reducing the amount of slush hydrogen fed through the engine (the reactor exhaust heats this up; less slush hydrogen means a hotter, thinner exhaust, trading off absolute thrust for higher Isp).

That said, I'm extremely skeptical of the design of the fusion reactor used in the study. The underlying principles are sound, but they're proposing to get well past ignition with a fuel 50 times harder to use than D+T in a reactor vastly smaller than ITER. If this is a realistic reactor design, build it on earth and run a power plant with it. On the other hand, the Longshot design made very optimistic assumptions about its reactor too, and the Daedalus design made *extremely* optimistic assumptions, so par for the course. --Christopher Thomas (talk) 08:02, 2 August 2010 (UTC)

Yes, I too am skeptical about D³He fusion, certainly until we find a credible source of ³He. It is not clear that there is any reachable (ie, outside the gas giants) in the Solar System. The mass spectrometer on the Huygens probe of Titan in 2005 found no He at all, to a very low level. I think fission/electric will rule the Solar System in this century, with maybe D-T fusion in 50+ years pushing us out into the Oort Cloud, if we are lucky. The use of H slush for reaction mass makes some sense for Solar System travel at v<400 km/s, as it greatly reduces the power problem: high acceleration (>1e-3 g, in our context) and high exhaust velocity together imply huge exhaust power, which inevitably brings along with it tremendous waste heat and power-to-mass problems. This is for me the central knot to be solved for "classical" interstellar transportation systems. Some kind of completely non-thermal reaction propulsion is what we really need, but lacking that (? Does anyone know any promising non-thermal fusion reactions? Calling on Dr Pons...) high efficiency is the best we can do. So I think this is only "one small step" on the way. Wwheaton (talk) 19:53, 2 August 2010 (UTC)

Helium-3 is about 1 ppm in normal helium on Earth (though our supply is dominated by Helium-4 produced by alpha decays). If relatively small amounts are needed, it could certainly be extracted from industrial helium supplies. The moon has a higher concentration (He-4 is 28 ppm in the lunar crust, He-3 is 0.01-0.05 ppm, for a relative fraction of 300 ppm or better), so helium mining on the moon is occasionally proposed as a possible lunar industry. In the near term, the only thing we're going to be using is D+T, because dealing with the neutron flux is still easier than building a reactor with 50 times better confinement. If we needed large amounts of He-3, I imagine we'd do it by breeding tritium from lithium and letting it decay. If we have a reactor that good, though, I'd just use D+D and live with the extra neutrons. --Christopher Thomas (talk) 21:07, 2 August 2010 (UTC)

Yeah. My enthusiasm for strip mining the Moon is muted, but if that turns out to be the most practical route, then so be it. I'm very unclear about the real difficulties in moving to the more difficult D³He reaction on the 50-year time scale we are probably faced with here. Given the lack of the kind of limitations on linear size and mass we are used to on Earth in the space environment, there may be unanticipated advantages in going to much larger designs, assuming the use of extra-terrestrial materials frees us from launch cost limitations to LEO. If the Discovery II design were scaled up 100× to the size of today's large crude carriers, how would that affect the containment and neutron problems? I'd also like to know more about the ultimate power-to-mass prospects for nuclear fission in space, especially coupled to ion drives, as that will determine the accelerations achievable, which is in turn directly related to trip times. (Bearing in mind the rule of thumb, 1g for 1 year -> rapidity 1, ~1c, so eg 5e-3g for 40 years gets us to 0.2c, etc, or an average of 0.1c for an 80 year start-stop trip.) Although I believe fission is not energetically up to the interstellar objective, we will need the much higher exhaust velocities that ion drives offer, compared to the <500 km/s Discovery II example. Wwheaton (talk) 16:12, 3 August 2010 (UTC)

Regarding fusion, it turns out that you can scale up a magnetic bottle to be as large as you like when gravity is removed from consideration. The sheet current needed for a given magnetic field stays constant (per metre), which means the magnetic pressure from a given field strength stays constant. Increasing the dimensions by a factor of x means tension in the structure goes up by a factor of x (due to increased radius of curvature), but the wall thickness goes up by x as well, so the tensile strength needed stays the same. For the types of reactor we've been trying to build, energy loss is dominated by plasma drift, not by radiative losses, so making a vessel x times larger lets you hold on to the plasma for x times longer, so scaling things up does help with respect to achieving the Lawson criterion. It's just more attractive to try to boost the plasma density instead, as reaction rate goes up as the square of that (that's why present research has been on improving understanding of plasma turbulence modes and how to suppress them, and why the Discovery II design has a density spike in the middle of the toroid). Boosting field strength is also attractive (density is proportional to the square of field strength, so reaction rate is proportional to the fourth power of it), but we're near the limits of what we can achieve in a reactor environment, barring discovery of better superconductors.

With regards to using large reactors for space travel, the problem is that reactor mass scales as the cube of its size but ability to dump heat scales as the square of its size. Even with a skeletal frame, like the Discovery II proposal, the parts that are actually solid get a heat flux that goes up with size. You get something similar with the neutron flux; a large reactor could absorb all of them, by using a few metres of graphite shielding (it slowly transmutes to C-13, which is a good neutron reflector, and then C-14, which is an excellent neutron reflector and stable enough for our purposes). However, the neutron flux and its associated heat load also go up with size for the same reason.

With regards to fission, the limits to power _generation_ density are your ability to reject heat, and the power to weight ratio of your conversion turbines (I'm assuming that it's better to use turbines and have the generation ship's crew actively maintain them than to build much-less-efficient thermocouples and hope they last a thousand years). Turbine power to weight ratio is ridiculous (I ran numbers for military aircraft a while back, and they get something like 10 MW/kg of turbine). Heat rejection dominates. This is optimized by using a working fluid other than water, having the hot end be as hot as you can keep the core without structural damage (around 2500 K, limited by structure as the fuel melts around 3000 K) and the cold end as hot as you can tolerate efficiency-wise (higher temperature improves heat rejection but reduces Carnot efficiency). Putting the cold end at about 1200 K gives you at most 50% efficiency, but the ability to dump about 250 kW/m² radiatively (compared to about 1 kW at room temperature). What this translates to in terms of power to weight depends on the assumptions you make, but I'd say 10 kW/kg is a reasonable upper bound to power to weight ratio including radiators, piping, turbines, and reactor.

Power to weight ratio for the drive system depends on what system you use and what assumptions you make, but something like a VASIMR drive (which uses cyclotron-frequency RF to heat plasma in an open-ended magnetic bottle) would be efficient enough that the radiators are likely to be the limiting factor. If you can keep the RF drive circuits and coupling antenna in resonance (mostly by changing an antenna loading tuning element to cancel the effects of the plasma on the antenna's near-field behavior), then losses are minimal. Other electric drive options work too; this just has the virtue of being easy to stuff a lot of power through (not sure an ion drive or a Hall thruster can do that). --Christopher Thomas (talk) 18:33, 3 August 2010 (UTC)

Interstellar particles and collisions

I'd like to point out that the so-called "vacuum" of space still has particles (like hydrogen atoms, dust and other material) between the stars. Not only is there a probabibility of a collision causing mechanical damage or destruction to the spacecraft, it will be exposed to significant amounts of radiation colliding with particles of atomic scale when traveling at extreme (near lightspeed) velocity. While modern electronics is somewhat resistant to radiation, it can only take so much. So more discussion of this would be prudent in this article. --71.245.164.83 (talk) 23:08, 5 October 2010 (UTC)

For speeds substantially below the speed of light (pretty much all of the design studies performed to date), interstellar gas doesn't behave like high-energy radiation (or at least, doesn't contribute a dose that's significant compared to what the craft gets from ordinary cosmic radiation). I agree that for speeds very close to that of light, this would become a significant issue.

The main difficulty that comes from insterstellar dust and gas is slow ablation (wearing) of the craft. Despite being called "dust", most components of the interstellar medium are extremely small; small enough to cause wear, not cratering, on impact. The density is also extremely low; on average in the galaxy it's about one atom per cubic centimeter, and in our local region it's a tenth that (for at least several dozen light years around). For any craft traveling within this region, we'd be plowing through about 1e+21 atoms per light year per square metre of forward-facing area, or about 1 mg of matter (given that it's mostly hydrogen). Even at high speeds, that won't cause catastrophic plasma-etching as long as there's some form of shielding in the way.

The most clever way I heard of dealing with the problem was to spray a fine mist of oil droplets forwards from the ship. The droplets are given a small electric charge, and a magnetic field is used to hold the cloud in position. The oil cloud would be heated and slowly degraded by the interstellar medium, but the craft would be protected. More conventional approaches just use bulk shielding of some kind. --Christopher Thomas (talk) 03:33, 6 October 2010 (UTC)

Good response Chris. Appreciate the response and keep it coming - it adds credibility. Material shielding (oil or whatever) is extra mass to carry, but a neat idea. Not all spaceborne particles will be of atomic hydrogen though. More research is probably needed. Just pointing out one of the risks. --71.245.164.83 (talk) 23:37, 9 November 2010 (UTC)