The launch loop couples very inexpensive velocity from an electrically accelerated linear rotor, and transfers momentum to the Earth via ambits (turnarounds) at the end.
A typical circular Low Earth Orbit (LEO) has an altitude perhaps 422 kilometers above the 6378 km equator, a radius of 6800 km. The orbital velocity v is 7660 meters per second, and the kinetic energy (½ m v²) is 29.3 Megajoules per kilogram. The climb from the surface to altitude requires an additional 2.1 Megajoules per kilogram. At 100% efficiency, that's 31.4 MJ/kg, equivalent to 8.7 kilowatt hours per kilogram. Here in hydropowered Oregon, my November 2018 electric bill was $0.122 per kilowatt hour, so the total energy is $1.06 per kilogram at perfect efficiency. Perfect efficiency isn't possible, but a launch loop can come close.
Rockets are nowhere near close; almost all the energy ends up spewing heated exhaust out the back, because rockets must be self-contained, and couple their momentum only to propellant, not the Earth itself. In 2018, SpaceX will launch 63,800 kg to LEO for $90M, or $1410 per kilogram, more than 1000 times more expensive.
But the launch loop is a 3.5 megatonne per year system, and will cost billions to build a minimum of three (for redundancy). The typical aerospace "learning curve" is a 15% cost reduction for a 2x increase in volume; a volume increase from 200 tonnes to 4 million tonnes per year might lower the SpaceX price by a factor of 10, to $140 per kilogram. The loop is high risk and capital intensive, so its "burdened cost" might be $3 per kilogram. Still, a cost savings of 50x may be possible.
If the launch loop is powered by space solar power, it can launch its own power supply, greatly reducing costs and producing very little pollution. Not zero; re-entering vehicles will generate ozone-destroying nitrous oxides . . . but not nearly as much as rocket launches.
Stability, Electronics, and Computation
The track/rotor spacing of a launch loop is nonlinear and unstable. This can be corrected with accurate measurement and electronic control. Viewed as an electronic problem, an 14,000 meter-per-second rotor moves at 14 micrometers per nanosecond. In 2018, An nVidia V100 GPU can perform 14,000 single-precision multiply-adds per nanosecond. Typical instability doubling times are more than 10 microseconds for the loop, and control lengths are on the order of a meter, so "one GPU" of computing could stabilize the entire launch loop.
In reality, the controllers will be small, micropower, multiply-redundant and distributed along the entire length. They will scavenge power from taps along the rotor, and may distribute power and control signals with gram-per-kilometer, 20-watt-capable optical fiber. The technology of 2018 is more than capable of controlling a "slowly moving" launch loop.
. Launch loops are not mobile; they can only launch eastward to velocities well below rotor speed.
- A Hohmann transfer orbit to Mars requires a launch velocity of 11.48 km/s; adding velocity for drag, and subtracting the Earth's rotation velocity, that is "only" 11.1 km/s loop launch velocity. The asteroids and outer planets will need higher velocities.
. Launch loops store a vast amount of energy,
From Now to Then
With current launch rates less than 10,000 tonnes to orbit per year,