If cost were no object: Interstellar travel

Factual Questions
41

Given that 80% of the initial energetic yield is neutrons (a fast neutron at 14.1 MeV) you are going to need a hell of a lot; not just of mass but density in order to get an aggregate neutron cross-section to capture a significant portion of that energy. And of the other 3.5 MeV of alpha particles, a fair portion of that is going to end up radiating x-rays due to Bremsstrahlung losses. Again, this requires a material opaque to x-rays to absorb the energy, thermalize it, and then convert the thermal energy to directional kinetic energy. Bigger devices and more propellant materials helps, certainly, but the assumption that the entire radiative output of a nuclear weapon will be converted to useable kinetic energy is not even remotely feasible.

Why would you think that extracting kinetic energy from this unconfined and radiating system would be anywhere close to the Carnot efficiency? The Carnot cycle assumes isothermal expansion and compression, and isentropic (no loss) heating and cooling. Even setting aside how much of the energetic yield is not converted to kinetic energy, the thermal efficiency of a system (from the heat radiated to the pusher plate, back to the cooling plasma, and to the 2.7 K microwave background) would be enormous, and because the gas is unconfined, you only extract as much energy as a the pressure equilibrium (when the pressure of the expanding gas cloud equals the inertial forces on the vehicle). This is literally like propelling your car by throwing sticks of dynamic out the back window and letting the impulse push you along.

That depends on the temperature profile of the cloud, the opacity of the material, the mass composition (lighter is better as it gives a higher effective velocity), and its thermal radiative characteristics (i.e. how emissive it is). Scaling up the bomblet size improves the efficiency since the ratio of volume to surface area of a spherical bomblet cloud scales by r, but there will still be significant losses even within the cloud between the gas interacting directly with the pusher plate and that expanding in other directions. The energy efficiency of such a system is quite low and the conversion of energetic yield to kinetic energy with momentum in the vehicle axial direction which limits the amount of total impulse available, which is another reason that the assumption of complete conversion of energetic yield to useable impulse is not a basis for a realistic estimate.

A million tons may only a modestly sized dam, but in terms of a mobile structure is is enormously larger than anything ever constructed on Earth. The largest moving structure I can think of is an aircraft super carrier, which is somewhat less than 100,000 tons. Consider the task of having ship the equivalent of ten supercarriers up to orbit and then assemble them, and you can see how absurd the task would be, not withstanding how problematic it would be to design it such that major structural pieces such as the pistons or pusher plate could be flown in payload-sized chunks.

The current launch capacity of heavy lift launchers (we’ll call heavy lift 9 tons to Low Earth Orbit just so we can include the Atlas V) is around 20 launches a year. There are currently zero launchers which can lift 100 tons to LEO. The American SLS, if it ever launches, is advertised to carry 70 to 130 tons to LEO, with an assumed launch rate of around 2 per year. The next largest is the Falcon Heavy, which is advertised to carry 53 tons to LEO. Even if we assumed that we could scale up a vehicle like the Falcon Heavy to 100 tons capacity and launch two a week without loss or stoppage (which is ridiculously optimistic) it would still take a century just to lift all of that mass to orbit, notwithstanding everything necessary to assemble it.

Dyson et al actually planned on constructing the Orion on the surface of the Earth and flying it into orbit using pulse propulsion. Setting aside the issue of fallout, the failure modes should the vehicle lose propulsion are unacceptably catastrophic; there is certainly no way to terminate that large of a vehicle into marginally hazardous chunks, and given the payload of a massive number of nuclear devices (even if it just carries enough to get to orbit with the intent of fully loading it later) I doubt the hazards would be palatable even if it didn’t overfly any populated location. Realistically, you’d have to construct such a vehicle in orbit, fabricating all major structural components from materials sourced in space. This alone is a massive infrastructure project well beyond current technical or logistical capability.

And the answer is no; not even if you put the entire world population and resources into it, it is not feasible using existing technology and logistics, nor would you be able to achieve speeds sufficient to transit interstellar space to even the nearest planetary system in a human lifetime. The technical viability of pulse propulsion is demonstrable, but attaining the performance required to travel to another star–much less in a human lifetime–is far beyond a reasonable estimation.

Let’s look at the mass ratios involved, just using specific impulse numbers. At I[SUB]sp[/SUB] = 5000 s, getting to 0.01c would require around 3x10[SUP]26[/SUP] kg of propellant for every kg of payload. Decelerating back down (presuming the destination system is relatively stationary compared to Sol) would require about 9x10[SUP]52[/SUP] kg of propellant for every kg of payload. If we improve the performance to 20,000 s (a realistic limit for a spacecraft we could actually construct) the ratios are only 4.2x10[SUP]6[/SUP]:1 and 1.7x10[SUP]6[/SUP]:1, respectively. Only at around 80,000 s do the numbers start to get remotely reasonable (45:1 and around 2000:1). Even at 100,000 seconds, you are looking at 21:1 and 440:1. If we up the desired speed to 0.03c, the requirements jump to 8x10[SUP]19[/SUP]:1 and 7x10[SUP]39[/SUP]:1 for I[SUB]sp[/SUB] = 20,000 s, and 9600:1 and 93x10[SUP]6[/SUP]:1 for 100,000 s, respectively. That still gets us to Alpha Centuri–the nearest star system, albeit one that is unlikely to have a planetary environment suitable for life–in about 140 years, or more than twice the average human lifespan. Notwithstanding the issues of maintaining a habitat and provisions for that duration, just keeping a working crew operating across roughly six generations is implausible.

So, no, crewed interstellar travel isn’t feasible using any extant propulsion technology, even assuming no practical limits on labor or resources.

Stranger

42

I want to point out that reaction mass is not required, but rather reaction momentum. Shooting matter at high velocity is the typical way of generating momentum, but one could also shoot photons or other fundamental bosons.

Also note that a matter-beam could also be used to generate momentum. For example, a process to shoot a collimated beam(s) of electron-positron pairs could serve as an engine with no reaction mass consumed.

Energy is the limiting factor instead of the amount of matter carried. And energy can be had for “free” via photo-electric collectors.

Here’s what I envision as an interstellar craft. A habitat shielded by water and other raw resources. Large panels that both collect energy from solar photons (a few thousand kelvins) and dump heat into the cosmic background (a few kelvins). Engines that emit collimated beams of photons and/or matter-anti-matter pairs. The power density is going to be very low, net force is going to be extremely low. But the friction of the interstellar medium is even smaller, so the craft can eventually reach any speed.

43

That depends on your threshold of acceptability. If the alternative is to be stepped on by a two-trunked elephant, then the risks might be considered acceptable after all.

Pleonast, “interstellar” and “solar power” don’t really go together very well. You’re going to need some other energy source.

44

That depends on if you are going to take the sun with you :smiley:

45

What comes to mind is E=mc^2 and they are actually the same, so if you are shooting photons, electrons, positrons or tachyon out your rear end you are losing mass.

Now if you can come up with negative mass, now we are talking about interstellar travel

46

Practically speaking, of course not. The power density is too low, we’d need panels much bigger than we can currently assemble.

But in terms of what is possible within the laws of physics, solar radiation is out there in interstellar space with thermodynamically available energy. It could be used as an energy source with appropriate technology.

Yes, but in a very different way than how rockets squirt fuel. It is laws-of-physics possible to collect photons, do useful work with the energy gained (including providing a continuous delta-v) and dump the waste heat back to the cosmic background. All without any long-term loss of the craft’s mass.

47

Fair enough. I’m finding it a little hard to find specs, but the impression I get is that even fairly thin (on the order of a centimeter) boron shields (of boron carbide, etc.) are fantastically efficient neutron absorbers.

The short, short answer is: why shouldn’t it? Of course you’ve given some good answers already, but the point is that there are a finite number of reasons why good efficiency can’t be achieved. To me at least, it’s more enlightening to figure out what’s theoretically achievable and work backwards, instead of starting from zero and working forwards.

The less short answer is that rockets–including, I think, NPP–have certain advantages that more well-known systems don’t. I just ran the numbers and an SSME has 67% Carnot efficiency, and 60% overall efficiency. That’s fantastic! Only the very best heat engines exceed these numbers. And that’s despite rockets being only marginally more “contained” than a NPP system.

I’d have to run more numbers to be sure, but I suspect a big reason is that rockets skip the compression steps, which are ordinarily a net loss. Someone else already paid for those steps when they condensed the liquid oxygen and such, so it doesn’t show up in the final efficiency figure.

And I think the same is true of NPP. The only steps to worry about are the heating and expansion. Correct me if I’m wrong here, but it appears that the heating is mostly icochoric (the mass vaporizes before it gets much of a chance to expand), and after that isentropic. I’m ignoring radiative losses for the moment, obviously, but according to my understanding this is as good as a Carnot cycle (i.e., they’re reversible steps).

If that’s true–and maybe it isn’t–then radiation is the main efficiency loss. I guess that’s just something I need to research more.

The shock absorber could be tuned to take this into account. The last part of the travel could be “geared down” so that it moves more readily as the pressure decreases.

And rocket engines are literally like shooting a fire extinguisher out the back. You know full well that it’s not a valid comparison; cars are efficient because they have the entire Earth as reaction mass. Rockets of all kinds have to balance mass efficiency vs. energy efficiency, and NPP is like all other high-Isp rockets in that it has absurdly low energy efficiency (in terms of converting energy to useful delta-V).

Okay. I guess I’m still interested in a quantitative analysis. The paper you linked to was interesting but didn’t contain these details.

Sure, but no one ever said it did. It’s just that Dyson assumed 25% efficiency for his lowball estimate and your number was 1.7% efficiency. You’ve given good qualitative reasons but not quantitative (I’m not saying that’s your responsibility or anything; just that that’s what interests me at this point).

I didn’t say it would be easy :). But orbital assembly does have certain advantages, such as no need for cranes.

I don’t think it’s even slightly optimistic. A Falcon Heavy has something like a 100 t dry mass. 20,000 launches means 2 million tons of machinery. In comparison, Toyota built about 15 million tons worth of cars last year, and rockets just aren’t that much more complex than a car per kg. Yes, obviously the current capacity would have to be scaled up by a very large factor, but it could be done.

Hell, Germany was building around 1,000 V-2’s a year under a mountain, with slave labor, in the middle of a war 70 years ago. Sure, a V-2 is a far cry from a Falcon Heavy, but it was still a real rocket and they were building a ton of them under dire conditions.

Sure. Though I tend to think of Orion Project hypotheticals in terms of: we’ve determined that a planet-killer asteroid will hit Earth in X years. Can we build something in time? In a situation like that, the downsides from fallout, etc. are irrelevant.

48

I was thinking in terms of a space elevator. By the time it was finished, the thing would be whipping around the Earth at a pretty good speed. Cut the tether, use a little chemical thrust to correct course, and you could be a good ways from the Earth before you started using nukes.

That’s sort of what I’m getting at: what is theoretically possible first, and practical second.

Given sufficient motivation, people often find ways of doing things that are “impossible.” The US went through 10 years of Depression because “impossible” for the government to do anything about it. Then we went to war, and suddenly there were enough resources to employ everyone. Imagine if all the resources that went into WWII went into something constructive, instead of destructive?

One attraction of nuclear power (and having zero science background is not helping me) is to my understanding you get a tremendous amount of power out of a small amount of matter. Which you need in order to overcome the “fuel is too heavy to push” problem.

That you’d need many more bombs than what the US and USSR ever produced doesn’t necessarily strike me as prohibitive. Only a small part of the cost of making the bombs (I suspect) went into making the bombs. They also had to make delivery systems - submarines, airplanes, sophisticated rockets, etc. And even all that put together was only a small part of the military budget. And then there were treaties that limited the number of bombs each side could have. I don’t know the cost of an individual bomb, but anyway, the point is that it wasn’t the cost per bomb that was prohibitive: there were other factors.

49

Solar power can be used to power the momentum-transfer method I mentioned earlier.

Solar powered lasers accelerate billions of thin, actively guided sails towards a much larger magnetic sail carried by the space probe. This has an advantage over simply accelerating a single sail by laser, in that the sails and the accelerating lasers can be smaller (according to Jordin Kare’s presentation, here)
http://www.niac.usra.edu/files/library/meetings/fellows/oct01/597Kare.pdF

There are other momentum transfer methods possible that also use solar power, including the use of particle beams and the direct use of lasers or masers on the probe itself.

Even antimatter could be considered as a form of solar powe r under certain circumstances, since one possible method of creating enough antimatter for use in interstellar craft would be to make it in solar-power arrays near the Sun, where energy is abundant.

50

Well, yes, that does help, but if you want to reach relativistic speeds (which is necessary to reach another star within a lifetime), it’s still not enough. See, the problem is that matter moving at relativistic speeds has an even more tremendous amount of energy for a small amount of matter.

51

See, I’ve never understood that. I mean, obviously matter moving at relativistic speed has tremendous energy. If a little grain of sand was moving toward me fast enough, it would have a ton of energy relative to me. (And I’d have a ton of energy relative to it, I suppose.) But how does the grain of sand know which one of us is moving at relativistic speed? I’m talking in terms of acceleration. If I try to push the grain a little faster, it’s hard. But if I try to push something else toward the grain, it’s easy. Yet the grain doesn’t know it’s the one traveling fast, does it? As far as it knows, it’s just sitting there, and other things are rushing toward it at relativistic speeds. So it should be just as hard for other things to accelerate toward the grain as it is to push the grain a little faster, right?

52

Don’t know from your post if you understand this but nuclear rocket engines typically mean that a nuclear reactor heats the fuel instead of chemical combustion of conventional rockets, we are not using the nuclear reaction directly as the propellent - also you are using fission. So you are just thermally heating material - so perhaps good for interplanatary travel, not for interstellar.

The Orion engine however does use nuclear propulsion directly as does some other fusion theoretical designs (Buzzard IIRC) and use fusion (same reaction as the stars). This way you get very high speed partials directly from the nuclear reaction as a propellent. This is the one with a chance of interstellar.

53

Energy depends on reference frame. A task might take a small amount of energy to accomplish in one reference frame, but the same task might require a very large amount of energy in a different reference frame. In either case, though, any observer can calculate how much energy it takes in any reference frame, and in any reference frame, the books balance (just in different ways).

54

Yeah, I know. That’s why I was talking specifically about Orion - it seems like it might have a chance of overcoming the carrying-the-fuel-with-you problem because it uses small bits of matter - atoms - to generate large amounts of energy.

55

So from the reference frame of the ship - Orion - it might take a small amount of energy, but from the reference frame of Earth it would take a lot?

56

This paper, if it’s speculations are correct, would mean that revolutionary advances in spaceflight could be possible in just the next few decades:

To summarize:[ul]
[li]calculations show that one type of quark matter might be stable, indeed technically more stable than ordinary nucleonic matter[/li][li]Such matter may not only be the Dark Matter ubiquitous throughout the universe, but would explain the imbalance between matter and antimatter.[/li][li]ultra-dense nuggets of such quark matter would probably serve as seeds for stars, planets and even asteroids.[/li][li]asteroids with rotation rates too high to be gravitationallly bound by ordinary mass may have ultra-dense quark nuggets at their cores, making exotic matter accessible in the near-term.[/li][li]This exotic matter would make possible the mass synthesis of antimatter, making interstellar flight feasible.[/li][/ul]

57

Thanks Lumpy, Antimatter drive would be great, and could be needed for reaching and colonizing the planets and moons (and also Pluto :p), but still interstellar travel would be difficult even with that.

Negative mass seems like the way to go at least for now.

58

Yeah. Imagine shooting a slow missile from a relativistic ship. From the ship, it took hardly any energy at all to accelerate the missile, but from the ground it took tremendous energy because it was already moving so fast.

The books balance because the process of firing the missile slowed the ship by a small amount (due to conservation of momentum). When you subtract out the energy that the ship lost, what remains is the ship-board figure.

A similar effect goes on with the upper stages of rockets. It sounds paradoxical, but these stages can often impart more energy to the payload than exists as chemical energy in the propellant. But there’s no free lunch–that energy comes from the kinetic energy of the propellant (i.e., the stage was already moving fast), put there by the earlier stages.

59

If money is no object, how about a megascale magnetic accelerator? One straight through the moon (~3500 km) should be long enough for a good kick. Of course, slowing down might be a bit of a problem…

60

A magnetic accelerator would need to be a lot longer than that to get to relativistic speeds, or even significant fractions of c.

Assume 3 gees of acceleration. The accelerator would need to be 375 billion kilometers long to get to 0.5c. That’s 2500 times the distance from the Earth to the Sun.