What are the ways to space besides a rocket from the ground?

Space is close: 100 km up, an hour’s drive if the road went straight up. Staying there is the hard part. To orbit, a spacecraft has to move sideways at about 7.8 km/s, fast enough that it falls around the Earth instead of into it. Height is a small share of the job: 20 km of height is worth about 0.6% of the energy of orbital speed.

A rocket has to carry the fuel for all that speed, and the fuel for carrying the fuel. In the NASA Glenn single-stage design the ideas below are measured against, 87.5% of the rocket at liftoff is propellant and only 2% is payload. That makes a rocket very sensitive to small gains: anything that saves a little speed, or lets the engine work a little better, goes straight into that 2%.

Neal Stephenson spent Blue Origin’s first years, from 1999 to the end of 2006, on studies of the alternatives. On Core Memory episode 94 he went through four of them with Ashlee Vance. Read the whole conversation in our Journal: Towers, slings, whips and space guns.

“The one that I still think somebody ought to build is the tall tower,” Stephenson said. Air pressure is the weight of the air above you, so at 20 km about 94.6% of the atmosphere is below the launch pad. Drag, it turns out, is the small part of the gain: in Geoffrey Landis’s NASA Glenn study the speed lost to drag falls by only about 1.4%. The big gain is the engine. In thin air it can wear a long vacuum-style bell from liftoff, and its average efficiency rises about 15%. The mass that reaches orbit grows by about 17%, and because the structure does not change, all of it goes to the payload: 2% becomes 4.1%, about twice as much.

The catch is the tower. Winds in the jet stream blow at 9 to 16 km, right through its middle, and a 2015 patent for a 20 km inflatable tower by Thoth Technology is as far as anyone has got.

SpinLaunch spins a projectile on an arm inside a vacuum chamber and lets go. Its planned full-size machine would release at up to 4,700 mph (2.1 km/s), about a quarter of orbital speed, and a small rocket would do the rest. Stephenson knew the idea as the Slingatron: “it’s a big gun, basically.”

The cost is the spin. Holding something on a circle takes a = v²/r: double the speed and the force quadruples. On an assumed 46 m arm, 2.1 km/s means about 9,800 g, close to the 10,000 g SpinLaunch designs its payloads for. Fighter pilots train for about 9 g. Then the projectile leaves the vacuum at Mach 6 into the thick air near the ground, about 23 atmospheres on its nose.

Longshot, in Oakland, fires payloads with pairs of gas tanks along the barrel that go off in turn behind the projectile. Its ~21 m test gun reached Mach 4.6. Its plan for orbit is a gun about 10 km long, and Ashlee Vance asked Stephenson about exactly that.

Length is the point. To reach a speed v in a barrel of length L with a steady push, the acceleration is a = v²/2L. The test gun averages about 6,000 g. Spread a near-orbital 7.5 km/s over 10 km and it is about 287 g, inside the 250–500 g Longshot is aiming for, which hardened electronics and rugged satellites can be built to take. People need about 3 g, which would take a barrel about 1,030 km long. And whatever leaves the muzzle at Mach 22 meets about 340 atmospheres of air: Stephenson’s warning that “all big gun systems have got problems … around what happens when you meet the atmosphere.”

“As a kind of almost tongue in cheek thing, I looked at gigantic steel bull whips,” Stephenson said, “which is still a topic I’m obsessed with.” A whip crack is a tiny sonic boom. A loop rolls down the whip, and as the whip thins the same energy is carried by less and less mass, so the loop speeds up. With a tip one tenth as thick as the handle, the end reaches about 32 times the loop’s starting speed, past the 343 m/s of sound.

That is nowhere near orbit, and Stephenson says so: giant whips or chain loops could not get there. But the rocket equation rewards any head start. A whole rocket starting at a bullwhip’s 379 m/s would need 93.3% of its liftoff mass as fuel instead of 94.0%, and that sliver raises its payload by about 36%.

Not because of physics. In those years, Stephenson said, the business of space was mostly launching huge, very expensive satellites to geostationary orbit, where the price of the rocket was “kind of a round off error.” Those satellites need insurance, “and you’re not gonna get insurance on a brand new launch … a radically new launch concept.” A delay of a few weeks is a few weeks of a satellite earning nothing. So Blue Origin chose rockets, and the radical ideas went to startups: SpinLaunch’s test flights in New Mexico, Longshot’s gun in Oakland, Green Launch’s hydrogen gun at Yuma. The tower is still waiting for someone to build it.

1. 1 Tower

   Open Tall Tower Launch and drag the tower to 20 km. The air above falls to 5.4% and the payload crates read ×2.05.

2. 2 Sling and gun

   In Spin Launch Lab, halve the release speed and watch the g-load drop to a quarter. In Space Gun Lab, stretch the barrel from 21 m to 10 km and watch 6,000 g become a few hundred.

3. 3 Whip

   In Whip Crack Lab, start from a rope and thin the tip to 1/10. The tip speed crosses 343 m/s and the whip cracks.

Each tool is a model with its assumptions written on it. The tower’s gains are interpolated from one single-stage design; the tower’s own structure, winds and cost are left out. The sling and the gun give steady or average loads, and real guns push hardest early, so their peaks are several times higher. None of them follows the payload through the air after it leaves, and both still need a small rocket to finish the orbit. The whip is an ideal tapered whip with no air drag, so its speeds are an upper bound. The films use short, credited excerpts of real footage; the numbers come from the models.