This section describes the various types of interstellar drive systems we have considered.
This is a relatively old idea first proposed by Robert Bussard. The idea was that a ship could scoop up interstellar hydrogen and use it for fuel. Since it wouldn't need to carry any of its fuel, it could accelerate to high speeds without concern for high fuel to weight ratio's.
Unfortunately, we don't really know what's in interstellar space, but we do know we are in a very thin part of it due to a recent (by galactic standards) supernova in our area. We might need a scoop thousands of kilometers across to fuel the motors for a decent sized ship (See below). Even if we could do that (which is highly doubtful), straight hydrogen is very hard to fuse, and doesn't fuse as quickly as we might need to run the ships motors.
All in all we have no real idea on how to make such a ship work; and if we could get it to work, we'd find it doesn't work very well in this part of the galaxy.
The problem with using a scoop, is their isn't much in interstellar space to scoop up. We found papers that proposed 1000 km diameter scoops that only weighed 200 tons. Assuming your moving at 1/3rd the speed of light (100,000 kilometers per second) with a scoop area of 1000 km across (pi*R^2=pi(50,000,000cm)^2 = 7.854E15 cm^2). You'd be scooping up the mass in 7.854 E25 cubic centimeters of space.
A big question is the composition of interstellar space. A classic assumption is that there is nothing but about 1 atom of hydrogen in a cubic centimeter of space. More recently, people guess it might be less than .054 atoms per cubic centimeter or as many as 10. Even more recently than that (say the last few months) it has been proposed that there may be a lot of long-chain carbon molecules in space. Perhaps 60-200 atoms / molecules. These small, dark, heavy molecules might be the missing 90-99% of the mass of the galaxy (euphemistically called "dark matter").
So far, no one really knows. This is unfortunate, because the composition of the interstellar medium makes a hell of a difference in the design of a RAIR-based starship. Since we don't know one way or the other, let's assume one atom per cubic centimeter at a proton mass of 1.673 E-27 Kg. At 0.333c, using the above design figures, our 1000 km in diameter scoop, scoops up a ram flow of 131.4 grams per second. That's about 345 tons a month. Given that the scoop weighs 200 tons (and you really want the mass at slower speeds) this really isn't very helpful. So this stardrive goes into the impractical bin.
A drive idea I came up with, and originally used as the assumed drive system for Explorer class starship design, was a multi-cycle Ram Augmented Interstellar Ramjet (RAIR). It would scoop up reaction mass from interstellar space like a pure ram scoop, but it would only use it as reaction mass, not fuel. It would accelerate this mass magnetically or electrostaticly using power from onboard fusion reactors. The scoop system could simultaneously scoop up fuel thrown ahead of the ship by a fixed launcher back in our solar system. So it wouldn't be limited in speed by the fuel it carried on-board.
If you could load a 1/4th light year track in space with enough fuel to keep the ship accelerating at 1g, the ship would (after 6 months) exit the track at half of light speed (0.5c). The ship could then switch to accelerating external scooped mass using power supplied by fuel stored on-board, or (more likely)coast to the target star. Assuming Alpha Centauri, in the later case it would coast for about eight years.
At low speed interplanetary trips, the drive would work like a conventional ion rocket or mass driver. Stored reaction mass would be feed into the electromagnetic or electrostatic accelerator core. (Unlike normal thermal rockets, an ion thruster works more efficiently with heavy atom ions. So I'll assume we are storing iron for reaction mass.) Power would come from fusion reactors running off stored fuel. Specific impulse varies depending on the exhaust velocity of the expended mass.
Now for the bad news - slowing down. We can't pre-load the deceleration course track with fuel at the target star because it would be virtually impossible across interstellar distances. Carrying enough fuel / reaction-mass to decelerate at the target star would be prohibitive unless the coast speed was kept low or another breaking force is found.
Getting out of the starsystem and back up to speed would also be a problem. If we decelerated using stored fuel. We know we can get back up to speed by refueling in the target star system. Slowing down when we get home would either require another magnetic brake, or use of a fuel launcher at earth.
This seems a clumsy and unreliable method, and given that this system is just a useless as a pure RamScoop (there still isn't enough mass out there to make it worth bothering to scoop up), this drive wouldn't work either.
Internally fueled Fusion Rocket.
A fusion powered rocket could cross interstellar distances, and is a near term enough technology to be considered likely for the mid 21st century. Unfortunately the amount of fuel it takes to get such a ship up to a usable speed (at least 1/5th of light speed is necessary, a 1/3rd or more is highly desirable.) can be far to much for a ship to carry. Possibly weighing hundreds to thousands of times as much as the rest of the ship.
The following table gives a breakdown of the fuel to ship mass ratios given fusion rockets of various specific impulses. Specific impulse is a standard way of measuring the performance of a rocket engine. For example a specific impulse of 1,000,000 (which gives an exhaust velocity of 10,000,000m/s); means that the engine gives 1,000,000 pounds of thrust, for one second, for every pound of fuel consumed. A specific impulse of 1,000,000 has long been a standard fusion engine performance number. (For comparison the best chemical engines have a specific impulse of 455.) Current designs might exceed 1,500,000, possibly more than 2,000,000. Which is fortunate since you'd need a specific impulse of over 2,000,000, with a 100 to 1 thrust to weight ratio, to be able to use this system to boost the ship So we would need to assume that fusion engines are developed, and advanced quite a bit before we could use them.
For example for a fusion rocket with a specific impulse of 1,000,000. If you wanted to use such an engine to accelerate a ship up to 1/6th the speed of light. The ship would need to carry 147 times its dry weight in fuel and reaction mass. If you want to get to 1/3rd the speed of light, it would need to carry 22,000 times its weight in fuel! Obviously no realistic ship could do this. Yet if we could build an engine with a specific impulse of 2,500,000 1/3rd of light speed becomes fairly reasonable.
| Specific impulse (exhaust velocity) | Speed 50,000,000 m/s (1/6 light speed) | Speed 100,000,000 m/s (1/3 light speed) |
| 2,500,000 sec (25,000,000m/s) | 7 to 1 mass ratio. | 55 to 1 mass ratio. |
| 2,000,000 sec (20,000,000m/s) | 12 to 1 mass ratio. | 148 to 1 mass ratio. |
| 1,500,000 sec (15,000,000m/s) | 27 to 1 mass ratio. | 785 to 1 mass ratio. |
| 1,000,000 sec (10,000,000m/s) | 147 to 1 mass ratio. | 22,000 to 1 mass ratio. |
| 500,000 sec (5,000,000m/s) | 22,000 to 1 mass ratio. | 500,000,000 to 1 mass ratio. |
Note: a specific impulse of 1,000,000 (A exhaust velocity of 10,000,000m/s) means that the engine gives 1,000,000 pounds of thrust, for one second, for every pound of fuel consumed. This has long been a standard fusion engine performance number. (For comparison the best chemical engines have a specific impulse of 455.) Current designs might exceed 1,500,000. (See write ups on Bussards Fusion reactor, and Bussards Plasma rocket) But none have a specific impulse of 2,000,000 or more, or get a 100 to 1 thrust to weight ratio. So
You start with a 1 billion ton fueled ship cluster, driven by 10 million tons of engines and support structure. Those engines are assumed powerful enough to push the whole mess with an acceleration rate of 10m/s. (A star ship that weighs as much as a thousand loaded super tankers, and is powerful enough to out accelerate a Corvette? Yeah right.)
When you burn off 95% of your weight in fuel. The ship cluster weighs 50 million tons, 20% of which is a first stage engine/structure that's WAY too powerful. You throw the first stage away and start a smaller second stage. It weighs about 400,000 tons (about as much as 4 aircraft carriers) and can push the 40,000,000 ton ship cluster at 10m/s. When you burn that down to 2,000,000 tons of cluster you throw that away that stage for a 70,000 ton ship with 5-10,000 tons of drive systems. Which can use the remaining 390,000 tons of fuel to get itself into the system.
| Stage | Total weight (In tons) | Thruster pack and stage structure. (In tons) |
| 1 | 1,000,000,000 | 10,000,000 |
| 2 | 40,000,000 | 400,000 |
| 3 | 2,000,000 | 70,000 |
So to get a 70,000 ton ship (with 5-10,000 tons of drive systems) into the target system. You need to launch out of this starsystem a one billion ton fueled ship. Even this assumes a 100 to 1 thrust to weight ration for a fusion drive systems (which is questionable), and once you get where your going, coming back is out (unless of course you scale the craft up accordingly). But it would give us huge fuel ratios for relativistic flight. So, in theory, a Multi stage fusion craft could get to the star. Assuming of course you can find a billion tons of fusion fuel, and a ship yard in space that can construct a ship the size of an asteroid! Which means in practice the ship is unbuildable.
Assuming you could build it, how fast could it get?
Assuming the engines had a Specific impulse of a 1,000,000 (Which as I said means it has an exhaust velocity of 10,000,000 meters per second), the speed at the point where you burn out the fuel for each stages is:
| Stage | Delta V per stage. Stages have 100 to 1 fuel ratio. | Delta V per stage. Stages have 20 to 1 fuel ratio. |
| 1 | 46,000,000 m/s (.15 C) | 30,000,000 m/s (.1 C) |
| 2 | 92,000,000 m/s (.31 C) | 60,000,000 m/s (.2 C) |
| 3 | 138,000,000 m/s (.46 C) | 90,000,000 m/s (.3 C) |
| 4 | 184,000,000 m/s (.6 C) | 120,000,000 m/s (.4 C) |
| 5 | 230,000,000 m/s (.7 C) | 150,000,000 m/s (.5 C) |
These numbers of course assume the ship has to carry the weight of its fuel. Obviously craft normally have to carry their fuel, but their are some ways around it.
We don't have to rely on the fuel nature left in interstellar space for us. A fuel launcher somewhere in our solar system, could throw the fuel out in front of where the ship is going to fly. The ship scoops up the fuel as its going along and uses it to fuel the engines. This has several advantages. The ships engines only need to accelerate the ship itself. (They don't even have to adjust for changing ship weights.) The fuel is accelerated up by the launcher, and the ship would only need a fraction of the fuel it would otherwise. This means the launcher system (who's power comes from unaccelerated fuel) takes up a large fraction of the load, and the ship saves a lot of energy.
Problems are that unless the ship is flying to a starsystem with a operating fuel launcher. It can't fly any faster then a speed it can decelerate from using its onboard fuel reserves. Also, this only works when your close enough to the launcher that it can accurately launch the fuel to you. Once your out of range, your stuck with the fuel in your tanks.
A fuel launcher based system (and the beamed power system listed below) has the advantage of eliminating the need for the ship to carry the heavy fuel (and power systems). Which improves the ships power to weight ratio significantly. But the systems are difficult to do, limit range, and don't seem to help us to slow down. (See details in Externally Fueled fusion Rocket)
Beamed power (or fuel launchers) have the advantage of eliminating the need for the ship to carry the heavy fuel (and power systems). That improves the ships power to weight ratio significantly. But the systems are difficult to do, limit range, and don't seem to help us to slow down.
Beamed power systems are most effective as microwave sail craft. But powered electromagnetic drives, or laser pumped drives are also possible.
The idea of a microwave sail is that you hang out a parachute like wire mesh. To the microwaves the mesh looks like a smooth reflective mirror. (Just like in the mesh radar dishes.) The microwaves bounce off the sail, pushing it, and the attached ship, forward. Just like a solar (or photon) sailer.
This idea has several advantages. Its efficient. The ship doesn't need to carry a complex drive system. You can even leave the main power system back home where all of the solar systems infrastructure can get at it for repairs or improvements. Unfortunately, it also has several disadvantages. You have to be able to hit the ship with the beam across interstellar distances. Since the ship will be light years away. You can't aim at it because your aim will be years out of date. The system also can't slow the ship down, since the pressure is straight away from our star system.
Robert Forward in his Roche World series of science fiction novels had a laser sail craft that got around this problem by having a two part sail. You drop the outer ring of the sail which, under robot control, tuned itself into a concave reflective mirror lens. The mirror would focus the light back toward the smaller, now reversed, inner sail. The outer sail, now free of the weight of the ship would be blown forward at fantastic accelerations, but the reflected light could decelerate the ship.
This system is simple in a theoretical sense, but impossible in a real world sense. The dropped reflection mirror would be under tremendous pressure from the power beam from earth. This would rapidly accelerate it away from the ship. Soon it would be so far away that time dilation would make it impossible for the mirror to aim at the ships retro mirror.
Fluctuations in the beam density would cause the mirror to ripple and dance like a kleennex in a jetwash. The mirror systems would need to be continuously surfing this light pressure to keep the mirror on the beam, and forcing the mirror back into precise curvature (with a mirror that could be hundreds to thousands of miles across), and aiming the reflected beam at the receding ship. Since light speed limitations mean it can't see where the ship is, it will have to calculate a probable position, based on its best guess of its beam precision and the ships deceleration mirror efficiency. Errors are inevitable, even ignoring random effects of erosion and system breakdowns; and any error will mean the mirror is missing the ship. Which would doom the crew.
Like the microwave sail above, a beam of microwaves is beamed to the ship. Unlike above the power is collected and turned into electricity to power the ships motors. Also unlike above, this can give reverse thrust. Unfortunately it might not be able to generate enough thrust to slow down. Much to our annoyance we found the sail effect on the collector would provide as much forward thrust as the engine could give going backwards, even if we assumed 100% efficiency.
As anyone whose ever seen an episode of Star Trek knows. Anti-matter and matter can be mixed to convert the total mass of both into tremendous amounts of energy. Pound for pound a Anti-matter / matter reaction releases well over a hundred times as much power as a fusion reaction. Unfortunately, though it releases more power; this power is harder to directly use to power the ship, and it is far more dangerous to handle. If we could synthesize the tens of thousands of tons of antimatter this would take. It would have the potential of exploding with a force of hundreds of millions of H-bombs.
We do not have the technology needed to synthesize, store, or ship anti matter on this scale, and don't seem likely to get it by 2050. Nor do we now know of any other way to convert matter completely to energy. If we did however, starships would suddenly become far simpler and more practical craft.