I finally got around to reading Larry Niven's 1971 novel A World Out of Time. It has a bunch of the really classic sf elements: bussard ramships, the distant future, uploads, distopia, genetic engineering, mobile planets.
The reason I haven't read this one before now is that it has a really ugly cover. This one really needs to be republished with a better cover.
Update 2006-09-17: Since I finished the book I'm been racking my brain trying to remember where I had read about the fusion candle. It was in a Schlock Mercenary strip from a couple of years back.
The candle is the most feasible method I've heard for moving planets. It involves improbable engineering, but doesn't involve impossible physics.
from pages 200-201 of A World Out of Time:
Two tubes, concentric, each a hundred miles long; the inner tube a mile wide, with thick walls if complex construction; the outer tube thinner and twice as wide. At one end, a bell-shaped rocket nozzle. At the other [...] Reworked military laser cannon, and vents, and a flared skirt, and thick stubby fins, there at the bottom end. Now temporary liquid hydrogen tanks were attached. Now the structure moved under its own power...it was a tremendous fusion motor...moving outward, circled by tiny ships...yeah.
Corbell said, "How do you move the Earth?"
"You are here to tell us that."
"Well it's a problem. A rocket motor won't do the job. You need something that can pull on the entire mass of the planet, nickel-iron core and crust and oceans and air, all at once. Where do you find a force like that?"
[...] "You move something else," Statholtz said. "The damage done by the rocket's thrust and by mistakes you might make will not kill anyone if nobody lives on the working body. Then the working body can be moved until the world falls toward it as a rock falls to the ground. What was the working body? Ganymede?"
"Uranus. Can you stop the light show at that picture?"
The lecture froze on an "artist's conception": a blurred, curved arc of Uranus's upper atmosphere. The motor looked tiny floating there. Corbell said, "You see? It's a double-walled tube, very strong under expansion shock. It floats vertical in the upper air. Vents at the bottom let in the air, which is hydrogen and methane and ammonia, hydrogen compounds, like the air that the sun burns. You fire laser cannon up along the axis of the motor, using a color hydrogen won't let through. You get a fusion explosion along the axis."
[...] "Okay. The hydrogen fusions in the middle of the motor---"
"---and the explosion goes out and up. It's hottest along the axis, cooler when it reaches the walls of the motor. The whole mass blasts out the top, through the flared end. It has to have an exhaust velocity way higher than Uranus's escape velocity. The motor goes smashing into deeper air. You see there's a kind of flared skirt at the bottom. The deep air builds up there at terrific pressure, stops the tube and blasts it back up. You fire it again."
"Elegant," says Statholtz.
"Yeah, Nobody's there to get killed. Control systems in orbit. The atmosphere is fuel and shock absorber both---and the planet is mostly atmosphere. Even when it's off the motor floats high for a while, because it's full of hot hydrogen compounds. If you let if cool off it sinks, of course, but you can bring it back up to high atmosphere by heating the tube with the laser, firing it almost to fusion."
Let's look at the numbers. Suppose that the acceleration is 1/1000 of a gee.
F = m a = (8.6832 × (10^25)kg) (0.01m/s^2) = 8.6832 × 10^23 Newtons
Assume the exhaust velocity averages 300,000 m/s (similar to a VASMIR)
mass flow = F / v = (8.6832 × 10^23 Newtons) /(300000 m/s) = 2.8944 × 10^18 kg / s
Which is insanely huge. We'd probably need a bigger exhaust velocity, but I can't imagine it being too much bigger.
New back-of-the-envelope calculations on this subject:
Let's say we want to calculate how much thrust you need to move Uranus to Earth's orbit using a standard Hohmann transfer orbit. Ler's only calculate how much thrust it would take to put Uranus into the transfer orbit, ignoring the second half of the problem---taking it out of that orbit.
radius at aphelion = ra = 2.9*(10^12) meters radius at perhelion = rp = 1.5*(10^11) meters gravatational constant = gc = 6.67*(10^(-11)) Mass of Sol = ms = 2.0*(10^30) Delta-Vee at aphelion = sqrt(gc*ms/ra) * (1-sqrt(2*rp/(ra+rp)))
bc to do this calculation:
#!/usr/bin/bc -ql ra=2.9*(10^12); rp=1.5*(10^11); gc=6.67*(10^(-11)); ms=2.0*(10^30); deltavee=sqrt(gc*ms/ra)*(1-sqrt(2*rp/(ra+rp))); print deltavee, "\n"; quit;
The answer was around 4700 m/s. Since the exaust velocity of our engine is 300,000 m/s, we need to use around 1/63rd of the total planetary mass to move it. (300000/4700 = 63).