Go to my Home page.
Go to my Joy of High Tech page.

The Joy of High Tech


Rodford Edmiston

Being the occasionally interesting ramblings of a major-league technophile.

Please note that while I am an engineer (BSCE) and do my research, I am not a professional in this field. Do not take anything here as gospel; check the facts I give. And if you find a mistake, please let me know about it.

Propelling us Towards the Future

     People who know a bit about technology (How does that saying about a little knowledge go again?) keep asking why NASA (or whoever) doesn't use better fuels than liquid oxygen (LOX) and liquid hydrogen (LH2). Something more exotic. Well, to start with, the proper term is "propellant," since you generally need more than just a fuel. (In the above example, hydrogen is the fuel and LOX the oxidizer.) The short answer is that we are already using the best propellants. Also, before about forty years ago liquid hydrogen was quite exotic enough for most propellant researchers and rocket engineers, thank you.

     The medium answer is that we are using the best propellants allowable by prudence, the law and the limits of our engineering technology. Which results in knowing nods, as the questioner decides that rockets don't burn nitroglycerine because we don't know how to use it safely yet.

     The long answer is that the laws of physics, chemistry and man being what they are, the propellants used today are as good as we are likely to have for the foreseeable future. (The preceding sentence is only the first part of the long answer. The rest is the article below.)

     Nitroglycerine isn't avoided because it is too powerful. In fact, if you count some types of solid rocket fuel, it isn't avoided at all. It isn't used to launch vehicles all the way to orbit because it isn't powerful enough. The best measure of a rocket's overall efficiency is the velocity of the exhaust (Ve) it produces. This one value takes into account a number of properties of the propellant and engine. (What happened to Specific Impulse, or Isp? This and the exhaust velocity basically measure the same thing. To get the Isp from the Ve values given here, divide by the acceleration of gravity, 9.81 meters per second per second. And, yes, Isp is measured in seconds, which is so weird to most people that the units are often dropped.) Under ideal conditions nitroglycerine would have an exhaust velocity of around 3880 meters per second. Naturally, in practice the value obtained would be somewhat less.

     There are double-base solid propellants (combining nitroglycerine with nitrocellulose or something similar, plus additives to regulate the burning rate and give the mix good physical characteristics) which contain as much as 50% nitroglycerine (though most are between 25% and 40%). These typically have exhaust velocities of around 1800 m/s. Compare this with the Space Shuttle Main Engines (SSME), which produce an exhaust velocity of around 4500 m/s. Even the theoretical maximum Ve for nitroglycerine is less than what we are already getting with LOX/LH2.

     Hydrogen and oxygen when combined release a lot of energy per kilogram, but there's more at work (pun intended) here than that. We want high exhaust velocity, which means the kinetic energy in the exhaust stream will be large. Kinetic energy equals one-half the mass multiplied times the square of the velocity. So you need lots of energy from the combustion (or decomposition for single-component propellants [monopropellants]), plus a low molecular weight. It is hard to get lower than H2O with any practical propellant.

     You may have noticed, however, that the Space Shuttle (more accurately, the Space Transportation System, of which the actual Shuttle is only a part) uses solid rocket boosters. These burn a combination of aluminum powder and ammonium perchlorate - the latter being the oxidizer - plus binders and plasticizers. This mix produces an exhaust velocity of around 2500 m/s. So why use these, instead of more or bigger SSME? Because thrust is a function of the momentum of the exhaust stream. Momentum equals the mass times the velocity. Which means that velocity isn't nearly as important if thrust is the priority. And at liftoff, with the vehicle at maximum mass and fighting gravity by climbing straight up, you need a lot of thrust.

     So there are conflicting requirements. For maximum efficiency a high exhaust velocity is needed, but that requires an exhaust stream with a low molecular weight. For maximum thrust the total mass needs to be high, which generally means a high molecular weight. And for both you need as much energy applied to your working medium (the exhaust) as you can manage.

     One reason for using multiple stages in a launch vehicle (though far from the only, and probably not even the most important, one) is to use different propellants in different portions of the ascent. The Saturn V burned liquid oxygen and kerosene in the first stage, LOX and liquid hydrogen in the second and third. The Space Shuttle burns LOX and LH2 in the main engines all the way from ground to orbit, but uses solid propellant boosters for the first part of the trip to help get off the launch pad.

     Knowing this helps explain why launch vehicles don't use mystery super fuels. There simply isn't anything practical with all the properties needed. This also explains why LOX and LH2 are likely to be the best propellants for the foreseeable future. However, someone with a greater knowledge of chemistry than those referred to at the beginning of this epistle will know that there are combinations of chemicals which yield greater chemical energy than LOX and LH2 with only slightly higher molecular weights in the exhaust. Burning hydrogen with fluorine releases more energy than combining hydrogen with oxygen, and fluorine is only one step to the right of oxygen on the periodic table, so the molecular weight of the product is only slightly higher. The exhaust velocity would be around 4700 m/s.

     However! Fluorine by itself is nasty stuff, capable of etching glass. Even if it could be handled safely, the exhaust would be hydrogen fluoride, which is very chemically active. If combined with water this makes hydrofluoric acid, a substance distressingly close to a universal solvent. The minor gain in performance isn't even worth the effort of preparing the environmental impact statement for a launch using fluorine and hydrogen.

     In rocket science, even a small change in propellant performance can have large effects. The goal of a launch is to put a payload into space. A tiny change in payload mass is amplified, because this means less fuel is needed, which means the tanks and fuel supply are smaller and less massive. This lower stack mass means you need even less fuel, which reduces the stack mass even more. An increase in fuel energy content would result in a similar cascade effect. As good as the SSME are - and they are very good - engineers are still working to improve their performance. A better fuel or oxidizer injector can mean both better mixing and less back pressure. The lower back pressure means you can use a smaller turbopump (or, more likely, reduce the existing unit's power so it lasts longer). And so on.

     As seen above, though, performance increases must be balanced against other considerations. LOX and LH2 are both cheap, and the tanks used to hold them in launch vehicles aren't all that expensive. A wonder propellant with twice the exhaust velocity of LOX/LH2 and enough thrust that solid boosters aren't needed, but which is 20 times as expensive as the cost of the total solid and liquid propellant load would not be used in the Shuttle. The LOX/kerosene combination is about as expensive as premium gasoline, but its Ve is only 3500 m/s. Using it in the upper stage(s) of a multi-stage rocket would make the lower stage(s) impractically large,  because the lower performance there would mean needing more of both kerosene and oxygen to get the same total impulse. (Note: LH2 has a very low density, and therefore needs larger tanks than kerosene for the same energy potential. However, hydrogen has more energy than does kerosene per unit of mass. If you can build lightweight cryogenic tanks (which we can) and can safely handle deep cryogenic propellants (which we can), hydrogen and its fuel tanks are less massive than kerosene and its tanks for the same energy.)

     One clever trick that has never quite worked is using ozone instead of oxygen. Energy is needed to cram an extra oxygen atom into the standard two-atom oxygen molecule. When you burn ozone you get the normal energy from combustion, plus that extra bit back. This additional energy results in a Ve (when burned with hydrogen) of around 4700 m/s, much like fluorine but without all that nasty active chemistry in the exhaust products. Problem is, ozone is metastable, and tends to spontaneously revert to standard oxygen, releasing the energy uncontrollably. The low temperature of liquid ozone slows this but doesn't completely eliminate it. And if one molecule decomposes, the energy released may cause two or three others to let go, starting a chain reaction which detonates your launch vehicle on the pad. Or your storage tanks in the tank farm.

     There are other combinations which theoretically produce better performance than LOX/LH2, and some of them even work. One family of exotic ingredients is known collectively as zip fuels, and involves various boron compounds. Adding small amounts of boron, lithium or both to LH2 can provide a significant increase in performance, producing exhaust velocities of over 5300 meters per second. However, boron and many of its compounds are very toxic, and are also pretty expensive. For commercial launches such tricks as adding boron or lithium to the fuel just aren't worth the extra cost and risk. Such propellants have been used for military rockets, but even there their application has been limited.

     So far all the propellants mentioned above have used chemical energy, with ozone adding a bit more from its decomposition. However, there are many ways of tapping the physical energy of certain materials. I mentioned ozone as a metastable substance. There are others. The physical properties of elements and compounds drive them towards certain favored configurations. Forcing them from those configurations requires energy. If they can be forced into a metastable condition (think of a ball in a shallow depression on top of a steep hill) this energy can then be released on demand. Just push the ball out of the depression and watch it roll down the hill. (That is, supply a small amount of triggering energy to start the reaction.)

     Monatomic hydrogen ("Single H" as Heinlein called it) has a theoretical Ve of 6550 meters per second. What happens is that we simply separate the two hydrogen atoms in a standard molecule, hold them apart until we want to boost, then let them come rushing back together. Lots of energy, plus a very low molecular weight in the exhaust. Problem is, no-one has ever been able to keep the stuff from recombining immediately after separation. Its metastable status is still theoretical.

     Spin-stabilized triplet helium (aka metastable helium) is another theoretically promising substance which we so far haven't had much luck with. The exhaust velocity would be nearly 31,000 meters per second. (That's higher than some fusion rockets.)

     There are many other metastable compounds which have promise, some still strictly theoretical and others realized only in the laboratory in minute amounts. Metallic hydrogen promises a solid monopropellant with vastly better performance than the best liquid rockets currently available, with a Ve of nearly 17,700 m/s. But how do you make metallic hydrogen? In industrial quantities?

     Yet another metastable propellant candidate is dodecahedral nitrogen, or N20, with a Ve of over 4900 m/s. This is also still a theoretical substance. A number of other nitrogen configurations have been studied, so far with no practical results for propellant use.

     Chemical and physical energy are dwarfed by nuclear energy. However, one of the restrictions I mentioned above was the laws of man. Nuclear power has been proposed for launch vehicles many times, and some development work done. One of the more spectacular concepts is the Orion. This is a cheap, high-performance method of propulsion which could send a crew of 400 to Pluto and back in two years, using a spaceship equipped with such luxuries as a swimming pool and a bowling alley. In the process it would use up a sizeable fraction of the world's weapons-grade plutonium and uranium. That's right; it uses explosions from fission/fusion bombs to push it through space.

     An Orion could be built in orbit, the components launched by chemical boosters. Since space is already full of radiation that consideration is greatly reduced, though you still wouldn't want to fly a vehicle or satellite through the exhaust before it dispersed. Once assembled an Orion could head for Pluto. Or anywhere else in the solar system. With an effective exhaust velocity of 735,700 m/s it isn't quite good enough for practical interstellar flight, but it makes a fine fliver for touring the local neighborhood. Too bad there's this international treaty forbidding the use of nuclear explosives in space. Even if you could get a waiver, how do you convince people to let you have several thousand mini-nukes, especially when you plan to haul them into orbit?

     If we ever get controlled fusion to work we can look forward to rockets with exhaust velocities of anywhere from 24,500 to 13,000,000 meters per second. The variation comes from the fact that there are various types of fusion and many ways of applying the energy released to propulsion. Even using the simplest method - exhausting the fusing plasma through a magnetic nozzle - thrust and performance can be balanced by adding extra hydrogen, or even water.

     Antimatter, ironically, would be easier to develop into a practical rocket... if we had enough of it to bother about. (Of course, if we had that much lying around we would want to get rid of it rather quickly, anyway. Just not too quickly.) Interestingly, for the best performance we would use about as much propellant as for a fusion rocket, because you would need to apply the energy released to a working medium, probably hydrogen. Your ship would therefore be as big as a fusion rocket have about the same total impulse. Oh, and the Ve for an ideal antimatter rocket would be around 299,792,500 meters per second.

     If we can't find a way around the light speed barrier the ultimate exhaust velocity is the speed of light, 299,792,500 m/s. Which is the velocity listed above for antimatter propulsion. That's because the result of mater/antimatter annihilation is mostly gamma rays, which, being electromagnetic energy, travel at the speed of light. The problem here is that photons don't have any mass, so the thrust of a photon rocket is understandably low. With an antimatter rocket there are ways to use the released energy to heat a working medium and use it to produce thrust. A pure photon rocket, though, is a different sort of beast altogether. Still, through the mysteries of relativity and quantum mechanics, photons do produce thrust, and a photon rocket will work. It's just that producing enough thrust to get up to speed anytime soon will require huge amounts of energy.

     So, you can see from this that there is great potential for improving rocket performance. You can also see that for at least the next few years things won't get much better.

     This material is Copyright 2002 Rodford Edmiston Smith. Anyone wishing to repost it must obtain permission from the author, who can be reached at: stickmaker@usa.net