(I'll give you fair warning: This blog entry is going to be only half-baked, a semi-stream-of-consciousness wandering about problems with chemical rockets, based on arguments I've gone through over the last couple of years. At some point in the future I'll likely clean it up, but for now, rambling is what you'll get.)
Rocketry is hard. Really hard. That is, to make a chemical rocket which successfully ignites and launches onto its intended path is a complicated piece of systems engineering. There are different kinds of "difficult" when you say that something is hard. In one (over-simplified) sense, making a space elevator is difficult mainly because of the extreme material requirements (i.e., something with a strength-to-weight ratio at least an order of magnitude better than Kevlar). Chemical rockets, on the other hand, are difficult because they require so many interdependent systems to be designed all at the same time. You have to iterate back and forth on your choice of and pressurization of fuel and oxidizer, nozzle cooling method, plumbing, heat transfer, thrust to weight ratio of the engine, et cetera, and et cetera, ad nauseum.
The big problem with chemical rockets is, well, that they rely on chemical combustion for propulsion. The basic idea of a rocket is to throw mass (your propellant) in one direction so that your rocket will move in the opposite direction. You want to throw the propellant with as high a velocity as possible, in order to be more efficient. You "throw" propellant generally by making it hot and then expanding it out a nozzle. Currently the most efficient way to generate heat in propellant is to combust it in a rapidly oxidizing environment — explode it, in other words.
There's just one minor problem with this design. Exploding chemicals are difficult to control. A rocket is generally 85-90% fuel and oxidizer, by mass, and so your rocket is pretty much a big pile of explosives, waiting for you to toss on a match. If everything goes right, you can control the flow of fuel and oxidizer, keep just barely on the safe side of your material limits, and direct your thrust in such a way as to put you into orbit. Any small leak of fuel, or just a tiny bit too much unexpected heat, or a control problem in your thrust, and you're suddenly S.O.L. (Shit Out of Luck, if you haven't seen the acronym before).
There's another problem with rockets: they don't get used very often. This lack of flight experience means that the design process has lots of guesswork in it, which must be compensated for by tons of testing. In the last few years, there have been about 60 rocket flights per year throughout the entire world. Compare that to the number of automobiles that have been sold, or even the number of airplanes that have been flown, and you'll begin to get some idea of the scale difference in testing and feedback that car manufacturers get, and what rocket builders get.
It's getting late, so I think I'll continue this thread some other time.
Posted by Tom Nugent at January 6, 2004 10:38 PM