CHEMICAL ROCKET LAUNCHER

Gas carried by the rocket is heated by a chemical reaction and expelled to provide thrust. At least 1 trillion (10¹²) U.S. dollars have been spent over nine centuries on rocket research. Unfortunately, rocket launchers remain expensive and prone to failure due to temperature extremes, enormous heat flux in the throat of the thrust chamber, severe vibration, low reliability of high-performance turbopumps, and the use of corrosive and flammable chemicals. Hot oxygen corrodes most structural materials in a matter of seconds. Refractory materials survive a few minutes. It takes extraordinary amount of research and experimentation to prevent meltdown and explosion of a newly designed rocket launcher. Most rocket launchers are expendable (disposable). The Space Shuttle is salvageable. Reusable rocket launchers do not exist. Cargo transported by rockets is called payload. The accent is on pay because it costs about $10,000 to launch 1 kilogram of cargo to a low Earth orbit. The ratio of cargo mass to the total mass of the rocket including its cargo and propellant is called payload fraction. Its value ranges from 6 percent for liquid propellant rockets to 0.2 percent for solid propellant rockets. The minimum mass of a chemical rocket launcher (for 1-ton cargo) is about 20 tons.

DETAILS

If we ignore gravity and aerodynamic drag, the final velocity of a rocket equals:

The total_mass includes everything: rocket engine, structural parts, propellant, and cargo. The dry_cargo_mass is the mass of the rocket engine, structural parts and cargo. The ratio of total_mass to dry_cargo_mass is called mass ratio (MR). According to the above formula, which is known as the rocket equation, a high velocity of exhaust gas is needed to launch massive cargo. Dividing the rocket launcher into several stages also helps, but it makes the staged launcher more complex than the single stage launcher.

Specific impulse describes propulsive efficiency of all contraptions which generate thrust by consuming fuel or rocket propellant. The specific impulse is defined as impulse produced by one kilogram of fuel or propellant. Metric units of the specific impulse are Ns/kg or m/s (meter per second). The specific impulse of a rocket is the same as its exhaust gas velocity, but the specific impulse of an air breathing engine is one order of magnitude greater. Most rocketeers express the specific impulse in seconds and calculate it by dividing the exhaust gas velocity by 9.8. All values of specific impulse given in this publication pertain to rockets immersed in a vacuum.

The maximum velocity of the exhaust gas is about twice its speed of sound:

U_max = A₀(2/(G-1))^0.5

where:

A₀ is the initial speed of sound of the exhaust gas

G is the ratio of specific heat at constant pressure to specific heat at constant volume

Failure of the Space Shuttle and other expensive rockets to reduce the cost of space transportation has persuaded NASA and private rocketeers to experiment with simple rocket launchers. They are often called 'big dumb rockets' or 'big dumb boosters.'

A rocket launcher must be reusable in order to be economical. This means that it must have liquid propellant engine with either regenerative cooling (plumbing circulating cold propellant), or radiative cooling, it must withstand high stress during atmospheric reentry, and it cannot use expensive, ablative shield protecting it from the heat during reentry. The high cost of the ablative reentry shield may be reduced or eliminated by:

1. Engine cluster.

2. The second stage of a rocket launcher has its engines on a side. When it reenters the atmosphere, it spins about its axis of symmetry, which is perpendicular to its trajectory. The spinning rocket is relatively cool because it radiates heat in all directions.

   

   Spinning rocket

3. A cryopump mounted on a satellite orbiting the Earth at the altitude of 160 km collects oxygen and nitrogen. At this altitude one cubic kilometer of the atmosphere has the mass of one kilogram. One half of the mass is oxygen, while the other half is nitrogen. The gasses are stored as cryogenic liquids. In a few months the mass of the collected gasses is the same as the mass of dry satellite. The gasses are liquefied and loaded into last stage of the rocket launcher. Liquid oxygen propellant slows down the last rocket stage before atmospheric reentry. During the reentry liquid nitrogen cools the last rocket stage. Electrodynamic tether or ion engine restores orbital energy of the satellite. (Similar idea called PROFAC was described in 1959 issue of JBIS.)

4. An orbital sling or an orbital coilgun slows down the last stage of the rocket launcher before reentry. The slow reentry is also better for the environment than the conventional, hot reentry which generates ozone destroying nitrogen oxides in the upper atmosphere.

There are eight types of chemical rockets:

1. Liquid propellant rocket burns a mixture of liquid fuel and liquid oxidizer, e.g., hydrogen and oxygen. Commercial quantities of hydrogen are made in a process called reforming: natural gas reacts with steam, producing hydrogen and carbon dioxide. The cost of making hydrogen is about 0.7 $/kg. The cost of liquefying and transporting hydrogen from the oil refinery to the user rises its cost to about 3 $/kg. Hydrogen, like chlorine, is a destroyer of the ozone layer. The cheapest and the least toxic fuels are methane, ethane, and propane. They are stored as liquids and cost about 0.4 $/kg. Liquid oxygen is the cheapest and the least toxic oxidizer. It costs only 0.05 $/kg. 98% hydrogen peroxide (H2O2) is not as energetic as liquid oxygen and is much more expensive (20 $/kg), but it can be stored at room temperature and it is better regenerative coolant. 70% hydrogen peroxide is much cheaper (1.4 $/kg) and easier to transport because it cannot explode, even when boiling at atmospheric pressure. Diluted hydrogen peroxide can be concentrated up to 90-98% by distillation and up to 100% by crystallization. Cheap grades of hydrogen peroxide are contaminated with hydrocarbons, which make them unstable; they spontaneously break down into water and oxygen. A mixture of the cheap hydrogen peroxide and a stabilizing compound does not break down, but the stabilizing compound contaminates the catalyst beds. Surfaces that are in contact with hydrogen peroxide must be passivated to prevent decomposition of the hydrogen peroxide. Kerosene fuel and its purified form known as RP-1 (Rocket Propellant-1) can also be stored at room temperature and they go well with the hydrogen peroxide. Unlike kerosene, RP-1 has uniform density, high coking temperature, and is free of sulfur. Coking produces tarry residue at elevated temperatures. Sulfur corrodes some metals and alloys. The overall density of the peroxide/kerosene combination is 1312 kg/m^3, better than overall density of liquid oxygen/methane, which is only 828 kg/m^3. On the other hand, the liquid oxygen/methane combination has 12% higher specific impulse and these propellants can be self pressurized, which means that their vapor pressure forces them into the rocket engine. The self pressurized rocket has simpler design and lower dry weight than a rocket using conventional pressurization system made of a helium tank and a flexible bladder separating the helium and the propellant. Small rocket engines are usually pressure-fed, which means that there is no pump and the pressure in the fuel tank and the oxidizer tank is high enough to deliver the propellants to the combustion chamber. The fuel and oxidizer are mixed and burned in the combustion chamber. To speed up the mixing, high-pressure fuel jets impinge on high-pressure oxidizer jets. The jets come out of holes drilled at acute angles in the injector and impinge on each other at a distance of about 6 hole diameters. (It is easy to break the bit while drilling the tiny holes at acute angles.) A misaligned jet spilling oxygen on the combustion chamber makes a hole in the chamber in less than one minute. The outermost injector holes inject only fuel to protect the combustion chamber with a layer of cold gas. This is called "curtain cooling." The exhaust gas rich in hydrocarbon fuel deposits protective layer of soot on the inner surface of the combustion chamber. The liquid propellant rocket using one kind fuel and one kind of oxidizer is called biprop rocket. The biprop rocket has a high specific impulse (3.0-5.3 km/s) but requires expensive engine. To reduce the size and cost of the engine, turbopumps are used to feed fuel and oxidizer at high pressure (2-40 MPa) to the combustion chamber. Turbopumps are the most expensive and the least durable parts of the rocket engine. Furthermore, the high pressure produced by the turbopumps necessitates the use of very efficient regenerative cooling and implies the use of weak propellant tanks which cannot survive reentry. Russians make the best turbopump-fed rocket engines. The thrust-to-weight ratio of their NK-33 engine is approximately 125. Cheap, simplified propellant pumps were proposed by Felix Godwin (Exploring the Solar System, Plenum Press 1960), Albert Sobey (U.S. patent 3,213,804), and Steve Harrington (Flometrics pistonless pump).

   

   Profile of liquid propellant rocket engine
   Thrust chamber consists of injector, combustion chamber, throat, and exhaust nozzle.
   Gas generator burns a mixture of fuel and oxidizer to generate hot gas which propels the turbine.

   

   Low pressure rocket engines made of iridium and rhenium can be cooled by radiation.

   Small rocket engines have many advantages over big engines. Engine development costs increase exponentially with thrust. Combustion instability is a common problem in prototypes of big, high-pressure, liquid propellant engines. The instability is caused by turbulence, which is determined by the Reynolds number.
   

   Reynolds number = R_e = V*D*S/N

   where:

   V = gas velocity

   D = diameter of the chamber

   S = gas density

   N = gas viscosity

   Gas viscosity is primarily a function of temperature. The impact of pressure is minor and the viscosity correction for pressure is less than 10% for up to 3.5 MPa. This means that the Reynolds number is proportional to the chamber's diameter and to the gas density, which is proportional to its pressure. The turbulence in small, low pressure engines is too small to disturb the protective layer of cool gas adjacent to the chamber's wall. This means that these engines are inherently more durable than the big, high-pressure engines. The high Reynolds number also damages large turbopumps because it is the main cause of cavitation and vibration. The most extreme engine instability results in pulsed flow of the propellant into the chamber. To prevent this extreme instability, the propellant must be pumped into the large engines at much higher pressure than the combustion chamber pressure. Small, pressure-fed engines do not need high tank pressure to prevent the extreme instability, so their tanks are relatively light-weight. Another reason for the superior thrust-to-weight ratio of small rocket engines is the Cube-Square Law which states that as scale is reduced, properties which are a function of volume (mass) will decrease faster than those which are a function of area (thrust and strength).

   The Cube-Square Law also applies to the entire rocket launcher. Structural stress during atmospheric reentry is so severe that it would destroy any rocket bigger than the Space Shuttle. The Space Shuttle is a compromise; it jettisons its external fuel tank to reduce its size and structural stress. Big, single-stage rocket launchers cannot return to the Earth, so they are neither reusable nor economical.

   Centimeter-size engines have as many advantages as disadvantages when compared with meter-size engines. The tiny engines weigh little even if walls of their combustion chambers are thick compared with their size. The thick walls conduct heat well, so they can be cooled by thermal contact with liquid oxygen tank instead of the expensive regenerative cooling. Failure of a few engines out of a few thousand is not a catastrophe, so poor quality engines can be used and reused many times. Tiny engines have tiny combustion chambers. A crude rule of thumb says that specific length of the combustion chamber must be at least 60 cm to ensure thorough mixing of fuel and oxidizer. The specific length L* is defined as the chamber volume divided by the throat area. If the specific length is less than 60 cm, the fuel and oxidizer must be very volatile and the injector holes must have diameter much smaller than one millimeter. Hydrocarbon fuel is not desirable because it may polymerize at high temperature and plug-up tiny fuel lines. Although miniature piston pumps can provide enough pressure, they are not nearly as efficient as the large turbopumps.

   Standard injector holes have cylindrical shape and diameter of about one millimeter. Smaller holes are desirable because they reduce volume of the combustion chamber and combustion instability. The best tool for making small injector holes is a powerful electron beam. The beam can make one hole having diameter of 50 µm in one millisecond. Other methods are at least one order of magnitude slower. Lasers are not suitable for making holes in aluminum and copper because these metals reflect the laser beams. Carbon dioxide lasers are reliable, inexpensive, and suitable for making holes having diameter larger than 50 µm. Excimer lasers can make holes as small as 5 µm, but they are expensive and less durable. The electron and laser beams can make tapered holes which improve the rate of propellant flow. Rounded hole inlets also improve the rate of propellant flow because they prevent cavitation. Mechanical drilling and EDM (electrical discharge machining) cannot make holes smaller than about 100 µm, and cannot make tapered holes.

   

   Injector holes

   Engine cluster is a multitude of decimeter-size engines carved in a monolithic slab of aluminum alloy. The slab is also the bottom wall of the liquid propellant tank and a heat sink during atmospheric reentry. The tank is integral part of the engine cluster. It holds liquefied oxygen or liquefied methane which are self pressurizing propellants. When pressure in the methane tank is too low, a pressure switch turns on a fan which churns the liquid methane inside the tank. The liquid absorbs heat permeating from the engine and vaporizes. Vapor pressure forces the liquid into a spiral channel carved in the slab. The liquid flowing in the channel cools the engine before entering the combustion chamber. Combustion chamber pressure is only 1 MPa because higher pressure would generate excessive heat flux in these poorly cooled engines. Expansion ratio of the engine cluster is about 10, its specific impulse is about 3 km/s, and its mass ratio is about 15 -- two times higher than the mass ratio of the big dumb boosters made of the same materials. Rocket cluster is a rocket launcher made of engine clusters stacked on top of a reusable hydrogen peroxide monopropellant rocket.

   

   Engine cluster without tank
   The cluster is divided into four independent segments.

   

   Void part of engine cluster segment
   Only the most important holes are shown - the real engine cluster has more holes to improve its cooling and to reduce its weight.

   The conventional engines are large, complex, hand-crafted, and made of pipes welded together. They cost about $7000/kg. The engine cluster is much cheaper because it can be made by a milling robot which carves the engine from a monolithic slab of aluminum alloy. (Similar method was used to make thrust chamber of the Agena rocket engine -- coolant passages were drilled in a monolithic slab of aluminum alloy.) The small size of the engines and low pressure of the exhaust gas reduces turbulence which is the main cause of premature failure of liquid rocket engines. The best material for the monolithic slab is aluminum 6061-T91 alloy. This alloy is machinable, weldable, and has thermal conductivity of 170 W/mK. Titanium alloys have greater specific strength (strength-to-weight ratio), but their thermal conductivity is too low for regeneratively cooled engines. The best drilling coolant for aluminum alloys is kerosene. Soft aluminum gums up ordinary drill bit flutes, so special drill bit is preferable. Aluminum alloy called Weldalite 049-T8 is the best material for the cryogenic tanks of the second stage because it is strong, light-weight, weldable, and has the same coefficient of thermal expansion as the aluminum engine. Its yield tensile strength is 690 MPa.

   The third stage of the rocket cluster is made of the aluminum engines and titanium tanks. Joining the aluminum engine with the titanium tank is difficult, but the titanium tank is better suited for the high temperature of the atmospheric reentry. A gasket made of Teflon or Kapton is wrapped around the aluminum engine before the engine and the tank are immersed in liquid neon. The engine is placed inside the tank and both parts are warmed up. A tight bond between the two parts forms as a result of different coefficient of thermal expansion of aluminum and titanium.

   Thrust produced by a pressure-fed rocket engine is so low that a long, pressure-fed rocket cannot lift itself off the ground. A cluster of short, pressure-fed rockets seems impracticable because it would generate too much aerodynamic drag during flight through the atmosphere. To reduce the drag, short engine clusters are stacked sideways and lifted to the altitude of about 30 kilometers by the reusable hydrogen peroxide monopropellant rocket. Each engine cluster has diameter of about one meter and length of about five meters -- one order of magnitude shorter than the length of conventional rocket launchers. The short rockets are relatively more sturdy than long ones. The engine clusters are strong enough to survive reentry, splashdown, and handling on a bobbing ship. During the reentry the engines are on the hot, fore side of the engine cluster. When the reentry begins the engines are used as heat sinks. When they reach maximum safe temperature, they are cooled by residual vapor flowing through them.

   

   Linear rocket cluster launcher
   The launcher is divided into square segments made of four engine clusters welded together. The segments are called quads. They fall off and splash down when they are empty. Third stage quad is protected during reentry by pivotal ablative shield (not shown).

   Although the hydrogen peroxide monopropellant rocket burns lots of expensive propellant, it is the best choice in the short term because its design is cheap and simple. In the long term the rocket should be replaced with a more efficient contraption: either a steamjet engine, or a helicopter.

   The helicopter used as the first stage of the rocket launcher (helicopter-rocket relay) is more reusable and slower than other first stage alternatives. It has two sets of propellers: small propellers generate thrust during flight through the troposphere, while big propellers generate thrust during flight through the stratosphere. The slow motion of the helicopter is perfectly suited for large, low density cargo, for example large telescope or large greenhouse. The launcher scales down very well because it is reusable and because Its atmospheric drag is negligible. The negligible atmospheric drag makes it possible to use very large exhaust nozzles which improve the expansion ratio and specific impulse. Last, but not least, the shape of the exhaust nozzles can be optimized for flight in the vacuum.

   

   Hexagonal rocket cluster launcher
   The hexagonal launcher is more compatible with the helicopter first stage because it is structurally stronger.

   The helicopter needs special engines that can operate at the altitude of 30 kilometers. There are three such engines:
   - Hydrogen peroxide monopropellant turbine.
   - Steamjet engine.
   - Electric motors.

   Electric motors powered by batteries are the best choice because they are cheap, reliable, safe, and easy to use. The motors and the batteries need a cooling system when they operate at high altitude. Magnesium hydride battery with Ni catalyst has the highest energy density but it is not yet mature technology. Li-ion batteries have energy density of only 534 kJ/kg, but they are very reliable and reusable. (They provide auxiliary power for my laptop computer.) The Li-ion batteries can be used as the power source despite their low energy density if used up batteries are discarded during the flight. It takes about 300 watts of helicopter power to lift 1 kg of weight. At the beginning of the flight the total weight of the batteries is about one half of the launcher weight. During 15 minutes of vertical flight the helicopter reaches its maximum altitude of 30 km, drops off nearly all its batteries on parachutes, and finally drops off the launcher. When the helicopter descends, most of its propellers (rotors) are used as wind turbines which provide power for the remaining propellers and charge the remaining batteries. The remaining batteries are used up during landing.

   If the cost and energy density of the rechargeable batteries do not improve in the future, the best source of power for the motors in the long term may be microwaves. Leik N. Myrabo pioneered the idea of microwave powered helicopters.

   Aluminum wires linking the motors with a high voltage generator standing on the ground would be expensive and difficult to use.

   Roton is a liquid propellant rocket which substitutes centrifugal force for the expensive turbopumps. Depending on design, either the entire rocket or its part rotates about vertical axis. It looks like the Hero engine, except that it has helicopter-like blades which provide lift during flight through the atmosphere. The centrifugal force is too weak to pressurize low density propellants, but it is sufficient to pressurize the hydrogen peroxide monopropellant.

   

   Roton profile

   Another unique design is the hydrazine catalytic decomposition rocket engine. It uses a single chemical: liquid hydrazine (N2H4). Hydrazine is very toxic and unstable at high temperatures. In the presence of a catalyst, hydrazine decomposes into nitrogen, ammonia, and hydrogen. The specific impulse in vacuum is 2.3 km/s. This reliable rocket engine controls the attitude of communications satellites and the roll of the upper stages of rocket launchers.

   A reusable hydrogen peroxide monopropellant rocket has similar design. 100% hydrogen peroxide monopropellant has density of 1450 kg/m^3, specific impulse at sea level of 1.6 km/s and exhaust gas temperature of 1285 K. 75% hydrogen peroxide monopropellant has density of 1330 kg/m^3, specific impulse at sea level of 1.15 km/s and exhaust gas temperature of 630 K.

2. Liquid rocket in tube is a liquid propellant rocket which flies inside a steel tube or a tunnel. The tube is filled with hydrogen gas to reduce friction between the tube, the rocket, and its exhaust gas. Liquid propellant tanks are surrounded by hot exhaust gas under high pressure. To reduce the rate of vaporization, the tanks are lined with thermal insulation. Liquid methane fuel and liquid oxygen are pressure-fed into the combustion chamber without the aid of turbopumps. Tank walls are thin and light-weight because pressure inside the tanks is only slightly higher than pressure outside the tanks.

   

   Liquid rocket in tube profile

3. Solid propellant rocket burns a solid block made of fuel, oxidizer, and binder (plastic or rubber). The block is called grain. Ammonium perchlorate oxidizer and other chlorine compounds are toxic, corrosive, and damage the ozone layer. Ammonium nitrate oxidizer is hygroscopic, but is usually more desirable, because it is safe, cheap, and smokeless. Solid propellant rocket is inexpensive, but has a low specific impulse (2-3 km/s), has to carry heavy casing, and cannot be throttled or stopped; it burns until all the grain is exhausted. When used in outer space, they may produce space junk in the form of micrometer-size aluminum oxide particles and centimeter-size slag.

   Mixing ingredients of the grain is dangerous because the mixing tool may scrape a solid surface and thus make a spark which ignites the grain. The liquid grain is cast into the rocket case and allowed to harden and cure. Extreme care must be taken during casting to ensure good bonding of the grain to the case wall and to avoid the formation of cracks and voids. The larger the rocket, the more susceptible it is to the formation of the cracks. A fast burning solid propellant may explode while it burns.

   

   Solid propellant rocket profile

4. Candle rocket is a solid propellant rocket without the heavy steel casing. The grain burns at one end like a candle rather than inside out. It slides under its own weight into the exhaust nozzle throat and burns there. Spiral grooves in the grain ensure uniform burning of the grain. The candle rocket cannot match the specific impulse of the solid propellant rocket unless it matches its combustion pressure, which is 5-10 MPa. The high pressure implies either high acceleration of the rocket, or a very tall rocket.

   

   Candle rocket profile

5. Solid rocket in tube is a solid propellant rocket which burns on the outside and flies inside a steel tube or a tunnel. Heavy casing of the ordinary solid propellant rocket is not needed because the tube holds the exhaust gas. The tube is filled with hydrogen gas to reduce friction between the tube, the rocket, and its exhaust gas.

   

   Solid rocket in tube profile

6. Hybrid rocket burns a mixture of solid fuel and liquid or gaseous oxidizer, usually synthetic rubber and oxygen. The rubber is perforated to ensure thorough mixing of the fuel and oxidizer. Hybrid rocket is exceptionally safe. It almost matches the high specific impulse of liquid propellant rocket, and requires only half the number of expensive turbopumps. Most designs forgo turbopumps; liquid oxygen is fed into the combustion chamber by tank pressure.

   

   Hybrid rocket profile

   Bruce Dunn post on sci.space.tech.
   Environmental Aeroscience Corporation.
   AeroTech RMS/Hybrid Reloadable Motor.
   SpaceDev Hybrid Rocket Program.

7. Inverse hybrid rocket burns a mixture of solid oxidizer and liquid or gaseous fuel. It is much less popular than hybrid rocket because the liquid fuel is highly flammable.

8. Pulse detonation rocket periodically detonates a mixture of liquid fuel and liquid oxidizer in a straight tube that has one end closed. Because the mixture is injected into the tube at a low pressure, turbopumps are not needed. Detonations do not bode well for the durability of this novel rocket. The specific impulse is about 10 percent higher than that of the liquid propellant rocket.

Steam rockets cannot be used as rocket launchers because their specific impulse is too low and their dry weight is too high.

Rapid manufacturing (also known as direct manufacturing) is an additive fabrication technology that makes complex plastic or metal parts automatically from data held in STL files. (STL files are made by CAD programs.) Metal parts produced by a sintering process are cheap, but they are porous and have poor corrosion resistance. Two technologies produce metal parts that are not porous and therefore resistant to corrosion and suitable for rocket engines. These technologies utilize laser or electron beam to melt metal powder feedstock.

The laser method is used by Optomec LENS-850 system. The system is slow (fabrication speed is about two cubic millimeters per second) and expensive. The maximum part size is 46cm x 46cm x 107cm.

Arcam EBM S12 electron beam system is faster than Optomec (about 17 cubic millimeters per second). The maximum part size is 20cm x 20cm x 18cm.

The best materials for regenerative rocket engines (aluminum and copper) reflect the laser beam (albedo up to 98%) rather than absorb it. The high albedo and poor energy efficiency of lasers (typically less than 10% of electric energy is converted to laser beam energy) strongly favor the Arcam electron beam system. A powerful electron beam is easier to generate and deflect than a powerful laser beam. Laser beams are deflected by moving parts which cannot match the scanning speed of the electron beam and require too much maintenance.

The best source of information about rapid manufacturing is the annual Wohlers Report ($390).

BIBLIOGRAPHY

K. K. Kuo and M. Sommerfield, (editors) Fundamentals of Solid-Propellant Combustion, AIAA, 1984.

Gregg Easterbrook, "Big Dumb Rockets," Newsweek, August 17, 1987.

Dieter K. Huzel and David H. Huang, Modern Engineering for Design of Liquid-Propellant Rocket Engines, AIAA, 1992, ISBN 1-56347-013-6.

Alain Davenas (editor) Solid Rocket Propulsion Technology, Pergamon Press, January 1993, ISBN 0080409997.

A. H. Epstein, et. al. "Micro-Heat Engines, Gas Turbines, and Rocket Engines: the MIT Microengine Project," AIAA 97-1773.

R.L. Bayt, A.A. Ayon, and K.S. Breuer, "A Performance Evaluation of MEMS-based Micronozzles," AIAA 97–3169, 33rd AIAA/ASME/SAE/ASEE Joint Propulsion Conference & Exhibit July 7-9, 1997, Seattle, WA.

M.N. Ross, J.R. Benbrook, W.R. Sheldon, P.F. Zittel, and D.L. McKenzie, "Observation of Stratospheric Ozone Depletion in Rocket Exhaust Plumes," Nature, Vol. 390, 1997a, pp. 62-64.

Oscar Biblarz, George Paul Sutton, Rocket Propulsion Elements, 7th edition, Wiley-Interscience, December 2000.

Alan Leo, "Pocket Rockets Pack a Punch," Technology Review, May 18, 2001.

Rocket Design by Tom Mueller:

Fundamentals of Propulsion (Microsoft PowerPoint file, 114 kB).

Rocket Propulsion Fundamentals (Microsoft PowerPoint file, 1047 kB).

Liquid Rocket Applications (Microsoft PowerPoint file, 723 kB).

TDK (liquid rocket engine performance program).

JANNAF (Joint Army-Navy-NASA-Air Force) Handbooks and Manuals.

CPIA (Chemical Propulsion Information Agency).

Innovative Rocket Designs.

Digital Micro-Propulsion research at California Institute of Technology.

Gregg Easterbrook, "Long Shot," The Atlantic Monthly, May 2003.

There are six newsgroups devoted to rocket launchers:

sci.space.tech

sci.space.policy

sci.space.moderated

sci.space.shuttle

sci.space.news

sci.space.history

History of rockets:

W. Von Braun, F. I. Ordway III, and D. Dooling, Space Travel. A History, Harper and Row, 1995.

Rocket history by Spaceline.

Rocket pioneers by John L. Sloop.

Brief history of rockets by NASA.

Early history of rocket launchers by Eugene M. Emme.

History of rocket launchers by Mark Wade.