!!c1QfXUgcGY0

This is not how spaceflight works

Gunnerkrigg Court, Chapter 37, Page 15.

Oh boy. That’s… man.

Okay. Let’s take this a panel at a time.

This is wrong in two ways, but I’ll cover only one here: Space launch is very heavily regulated. This is because, as everyone discovered in the 60s, there is very little difference between “a rocket that puts things in orbit” and “a rocket that carries nuclear weapons to cities on the other side of the globe.”

UN General Assembly resolution 1721 B (XVI) (Adopted in 1961) requires that member nations register space launches with the CPUOS.

In the United States, space launch systems are regulated under the Arms Export Control Act of 1976, and internationally by the ITAR. These control things you need for spaceflight, such as accurate gyroscopes, (Ring-laser or conventional, both of which no high-school student will have any hope of building from scratch) large amounts of chemicals that can be used to fuel a rocket, etc etc.

So we can see that either the United States does not exist in the Gunnerkrigg-verse, or that nuclear weapons, and thus ICBMs, were invented. Though, if there was no cold war, one would wonder why spaceflight would exist at all.

Space launches are also very closely tracked. The IR bloom of a launch is quite distinctive, and pretty much every nation with space-based MASINT capabilities are watching out for it.

Once in orbit, tracking a satellite is pretty trivially easy. Anyone with a DSLR and time can do it. People have been doing it for fun since 1994. Professionals do it professionally. The companion newsletter to SPACEWARN (for launches) is the Orbital Debris Quarterly News. (for, uh, orbital debris)

Nothing can slip into orbit, or hide, once it’s there. Anything that’s made of metal will reflect radar waves, and anything warmer than 2.725 kelvin will stand out clearly on infrared.

Here’s a blog post tracking some NRO satellites. They’re classified but not at all hidden.

There are some spacecraft designed to alter their orbits significantly, (like the X-37) to evade detection in the manner Microsat-5 is supposed to be able to do. This requires what space nerds call delta-v, and requires a big engine, as well as large fuel tanks. Supposedly the X-37 can do ~3.1km/s of delta-v. The X-37 also cost $493 million dollars to develop, and masses ~5,000kg. It’s orbited by a Atlas V 501, which masses 334 metric tonnes, and costs about $150 million a launch.

On-orbit movement is not cheap, or easy.

Speaking of things that are neither cheap nor easy, achieving orbit at all.

Low Earth Orbit, the easiest possible orbit to get to, so cheap that atmospheric drag will eventually deorbit anything flying that low, is 170 kilometres high, and 7.8 kilometres per second fast.

Scale check: 170km is about sixteen times higher than the cruising altitude of a 767.

7.8 kilometres per second is 17,000 miles per hour.

So, let’s do the math. Let’s assume that Microsat-5 is in LEO, and is literally named, i.e, is in the microsatellite mass class, between 10 and 100 kilograms. Since it’s been in LEO for a couple decades, we can safely assume it has some method of station-keeping. 10-100kg doesn’t give us a lot of reaction mass, so lets assume it uses an elecrodynamic tether, which maintains orbit by pushing against the Earth’s magnetic field.

We’ve got several options in the microsatellite weight class. The Juno II was a modified Jupiter MRBM rocket, converted to space launch after the Sputnik disaster. It’s 24 metres tall, and masses 55,110 kilograms. Its payload to LEO is 41 kilograms.

Speaking of Sputnik, why not use their launch vehicle? The Sputnik-PS (based on the first ICBM, the R-7 Semyorka) is 30m tall and 267,000kg heavy, but it can orbit a sturdy 500 kilograms of payload. 

Spaceflight: not cheap, not easy!

(Fun fact: R-7 derived space launch systems have been in continuous use for fifty five years. The Soyuz-FG is currently the world’s only manned spaceflight system, and it’s still pretty much the same rocket that launched Sputnik in 1957.)

There are additional challenges. Kennedy Space Center is not located in Florida because NASA is such a fan of the locals, mosquitos, or poisonous animals; and Guiana Space Center is not there because of the large aerospace manufacturing industry in French Guiana. 

The Earth spins once every 23 hours and 58 minutes. This means that a point on the equator moves at about 463m/s. In one of those neat tricks of physics that wouldn’t think would actually work, you can use the rotational velocity of the Earth to reduce the delta-v you need to put a satellite in orbit. Given the razor-thin margins of spaceflight, this can mean a difference of hundreds of kilograms of payload.

That’s the good news. The bad news is that the farther north you go, the more inclined the orbit has to be, and the less of a boost you get from the rotation of the Earth. This is why Russian rockets are generally huge: Baikonur Cosmodrome is 45 degrees north, while Kennedy is only 28. Gunnerkrigg Court appears to be surrounded by evergreen forest, which means they have to be fairly far north, which is going to make achieving orbit that much harder.

Let’s get back on track. That was all 60s technology, and we can do better. The Falcon 1e is the smallest, cheapest space launch system in the history of the Earth.

Its rated payload to 185km LEO is 1,010kg. To lift that much mass to orbit, you need a 46,760kg, 24 metre tall rocket, which costs 10.9 million dollars. $10,792 a kilogram, which means that Microsat-5 costs more than half a million dollars just to transport, not even counting how much it costs to build! This is with the best technology available, which required millions of dollars, years of development, and three full-scale launch failures.

Rockets are hard. Rockets are expensive. If you’ve been following the ongoing Armadillo Aerospace adventure, you might conclude that seven people, working full time, have spent ~$5 million and twelve years to build a rocket that crashes a lot. (Note: videos of rocket launches are loud.)

The rocket pictured in that panel can’t possibly make orbit without using magic. (Which it probably does.)

Hey! While I’m nitpicking, let’s complain about some other things that annoyed me.

Chapter 32: Bonus page.

Since the underwater dorms have a moon pool, we know that they’re pressurized. After spending 8 or 9 hours at depth, you’re gonna have to depressurize, or endure the amusing range of neurological symptoms associated with decompression sickness. Doing that once a day will be fun.

Chapter 33: Page ??

This is an odd thing for a machine to say.

From wikipedia:

In computability theory, a system of data-manipulation rules (such as an instruction set, a programming language, or a cellular automaton) is said to be Turing complete or computationally universal if and only if it can be used to simulate any single-taped Turing machine and thus in principle any computer. A classic example is the lambda calculus

In practice, this means all Turing-complete programming languages are functionally identical. Lisp is considered a vastly more expressive language than COBOL, (Amusingly, they’re about the same age) but there is no concept you can express in Lisp that you can’t do in COBOL. After all, this is how compilers work: you take a high-level human-readable programming language, and compile it to a lower level one, that the computer can understand. (Or, more commonly, via an intermediate language, like JVM bytecode of x86 assembler)

This has been taken to extremes in some computer science jokes: the brainfuck programming language, which has 8 characters: > < + - . , ] [

Those 8 characters only. No ASCII letters, or numbers. Hello world in brainfuck is ++++++++++[>+++++++>++++++++++>+++>+<<<<-]>++.>+.+++++++..+++.>++.<<+++++++++++++++.>.+++.------.--------.>+.>.

So there are possibly concepts which humans can’t understand, but there’s no Turing-complete language which an idea can’t be expressed in.

This would be one hell of a minor nitpick, except a fucking robot is saying it. I’d really hope a robot would understand computational universality.

8:54pm | author:
  1. 0player reblogged this from c1qfxugcgy0 and added:
    Disagreeing on the last point. Robot said “inefficient”, not “insufficient”, she may have as well meant “too low...
  2. pretenditsfunny reblogged this from c1qfxugcgy0
  3. c1qfxugcgy0 posted this