There are many different ways to accelerate spacecraft. Below is a summary of some of the more popular, proven technologies, followed by increasingly speculative methods.
The first rockets made by the Chinese used gunpowder, and until the 20th century all rockets used some form of solid or powdered propellant. Solid rockets are considered to be safe and reliable; these kinds of rockets have been used for so long that engineers have a very good understanding of them.
One of the many drawbacks to solid rocket systems is that once they've been ignited, they can't be turned off. That is why they aren't commonly used in satellites; satellites need rockets that can be fired multiple times (or pulsed.)
A hybrid propulsion system is composed of solid fuel and liquid or gas oxidizer, typically. These systems are superior to solid propulsion systems because of safety, throttling, restartability, stability, and environmental cleanliness. Hybrid systems are more complex than solids, though, and more expensive.
Most familiar combustible chemicals (gasoline, alcohol, wood, natural gas) require oxygen to ignite and burn, but there are some unique chemical compounds that burn by themselves - no oxygen required! This is because the combined chemicals, when energized, "reduce," or change in ionic state by reacting with each other, as opposed to an outside oxidant. The most commonly used monopropellant is hydrazine (N2H4), a chemical which is characterized as "strongly reducing". Another monopropellant is hydrogen peroxide, which when purified to 90% or higher is self combustable.
Rocket scientists long ago realized the usefulness of monopropellant chemicals for satellite propulsion. Because only one chemical is used, the system is very simple, and thus very cheap. Unfortunately, monopropellants are not nearly as efficient as the other propulsion technologies. Engineers choose monopropellant systems when the satellite's propulsion needs are not very great. If the propulsion system is going to be heavily used, such as on a geosynchronous satellite or an interplanetary spacecraft, other technologies are used.
Systems that use both a propellant and an oxidizer are called bipropellant systems. Bipropellant systems are more efficient than monopropellant systems, but they tend to be more complicated because of the extra hardware components needed to make sure the right amount of propellant gets mixed with the right amount of oxidizer (this is known as the mixture ratio.)
Dual mode propulsion
Dual mode propulsion systems combine the high efficiency of biprop with the reliability and simplicity of monoprop systems. Dual mode systems are either hydrazine/N2O4, or MMH/peroxide (the former is much more common). Typically, this system works as follows: During the initial high-impulse orbit-raising manuevers, the system operates in a bipropellant fashion, providing high thrust at high efficiency; when it arrives on orbit, it closes off either the fuel or oxidier, and conducts the remainder of its mission in a simple, predictable monopropellant fashion.
A resistojet can be thought of as a "space heater attached to a water hose". Of course, it's much more sophisticated than that! But it is a good analogy, because it emphasizes the two characteristic parts of the system: A source of heat produced by electricity through some sort of a resistor, and a source of (typically non-reactive) fluid.
Resistojets have been flown in space, and do well in situations where energy is much more plentiful than mass, and where propulsion efficiency needs to be reasonably high but low-thrust is acceptable.
Arcjets are a form of electric propulsion, whereby an electrical discharge (arc) is created in a flow of propellant (typically hydrazine or ammonia). This imparts additional energy to the propellant, so that one can extract more work out of each kilogram of propellant, at the expense of increased power consumption and (usually) higher cost. Also, the thrust levels available from typically used arcjet engines are very low compared with chemical engines.
Hall Effect Thrusters (HETs)
Also known simply as plasma thrusters, HETs use the Hall Effect to accelerate ions to produce thrust. A varient called stationary plasma thruster (SPT) has been used by the Russians for stationkeeping for many years and will be used on Western satellites soon.
One problem with HETs is that their plumes diverge quickly and thus can impinge on other spacecraft parts leading to thermal and contamination problems.
Example: Hall thrusters typically operate at over 50% thrust efficiency, provide specific impulse from 1200-1800 seconds, and thrust to power ratios of 50-70 mN/kW.
The plasma gas is typically xenon because of its high molecular weight and low ionization potential.
Of all the electric thrusters, ion engines have been the most seriously considered commercially and academically. Ion engines use accelerated beams of ions for propulsion.
Ion engines are best used for missions requiring very high delta-V's (the overall change in velocity, taken as a single value), interplanetary missions, for example. This is because the more performance required of the propulsion system, the faster a high efficiency system like an ion engine will pay off.
NASA has developed an ion engine called NSTAR for use in their interplanetary missions. Hughes has developed the XIPS ([[Xenon Ion Propulsion System]]) for performing stationkeeping on geosynchronous satellites.
Magnetoplasmadynamic (MPD) Thrusters
Thrusters that use the Lorentz force (a force exerted on charged particles by magnetic and electrical fields in combination) are called magnetoplasmadynamic (or MPD) thrusters. MPD thruster technology has been explored academically, but commercial interest has been low.
MPD thrusters can be run in a steady state fashion or in a pulsed mode.
Pulsed plasma thruster (PPT)
Pulsed plasma thrusters use an arc of electric current through a solid propellant (almost always teflon), to produce a quick and highly efficient and dependable burst of impulse. PPT's are great for attitude control, and for main propulsion on particularly small spacecraft (those in the hundred-kilogram or less category).
A gas, probably Hydrogen, is heated in a high temperature reactor, probably made of low-boron graphite. The heated gas expands through a nozzle, causing thrust. Prototypes of this form of propulsion were tested by the U.S. in the 1960s for use in Mars missions. The prototypes were developed by a program called "NERVA". They had specific impulses, (the measure of a rocket's goodness), substantially higher than any known chemical fuel. The advantage of this nuclear propulsion system was that it was a logical extension of both rocket and nuclear technologies, requiring no startling break-throughs in either. Real motors were built and tested at a Nevada test site before the program was discontinued.
Another variation, in a project called "Dumbo" planned to use a different, much thinner metallic reactor to achieve a higher thrust to weight ratio. The project developed some initial reactor designs, and appeared to be feasible.
NERVA could probably only be used for orbital transfer. Dumbo was designed to power heavy-lift launch vehicles.
Nuclear thermal energy is changed into electrical energy that is used to power one of the electrical propulsion technologies. So technically the powerplant is nuclear, not the propulsion system, but the terminology is standard. A number of heat-to-electricity schemes have been proposed.
One of the more practical schemes is a variant of a pebble-bed reactor. It would use a high mass-flow nitrogen coolant near normal atmospheric pressures. This leverages highly-developed conventional gas turbine technologies. The fuel for this reactor would be highly enriched, and encapsulated in low-boron graphite balls probably 5-10cm in diameter. The graphite serves to slow, or moderate, the neutrons.
This style of reactor can be designed to be inherently safe. As it heats, the graphite expands, separating the fuel and reducing the reactor's criticality. This property can simplify the operating controls to a single valve throttling the turbine. When closed, the reactor heats, but produces less power. When open, the reactor cools, but becomes more critical and produces more power.
The graphite encapsulation simplifies refueling and waste handling. Graphite is mechanically strong, and resists high temperatures. This reduces the risk of an unplanned release of radioactives.
Since this style of reactor produces high power without heavy castings to contain high pressures, it is well suited to spacecraft.
Deploys a large, lightweight sail which reflects solar radiation. The "light pressure" on the sail provides thrust by reflecting photons. Tilting the sail at an angle from the sun can produce a thrust component perpendicular to the angle between the sun and the spacecraft. Steering is usually with auxiliary vanes.
A number of demonstration projects have proven this method's feasibility. On a smaller scale, some unmanned spacecraft have been constructed with reflective panels that can be used to steer the spacecraft, to conserve fuel for maneuvering and attitude control. Solar collectors or sun shades can also serve as crude solar sails, and can help a spacecraft correct its orbit without using fuel.
It has also been proposed to use lasers to push solar sails. Given a sufficiently powerful laser and a large enough mirror to keep the laser focussed on the sail for long enough, a solar sail could be accelerated to a significant fraction of the speed of light. To do so, however, would require the engineering of massive, precisely-shaped optical mirrors (wider than the Earth for interstellar transport), incredibly powerful lasers, and more power for the lasers than humanity currently generates.
A mass driver would use an electromagnetic accelerator to move any mass up to high speeds. The bucket would then be decelerated, and the mass would keep going. If mounted on a planetary surface, mass driver could accelerate payloads to escape velocities. If mounted on a spacecraft, a mass driver could use any type of mass for reaction mass to move the spacecraft. This, or some variation, seems ideal for deep-space vehicles that scavenge reaction mass from found resources. Prototype mass drivers have been in existence since 1975, most constructed by the Space Studies Institute.
One possible drawback of the mass driver is that it has the potential to send reaction mass travelling at dangerously high relative speeds into useful orbits and traffic lanes.
Technologies requiring more engineering development:
A magnetic sail would deploy a large conductive loop to generate a magnetic field, and possibly auxiliary loops for steering or field-shaping. Magnetic sails are attractive because if superconducting, they have a lower weight-to-thrust ratio than solar sails.
There are three known operating modes.
Inside a planetary magnetopause, a magnetic sail can thrust against a planet's magnetic field, especially in an orbit that passed over the planet's magnetic poles.
Outside a magnetopause, the sail's magnetic field deflects the solar wind, the positively charged protons (ionized Hydrogen) continually emitted by the sun. Solar protons are far more massive than photons.
In interstellar spaceflight, outside the heliopause, a magenetic sail could act as a parachute, to decelerate a spacecraft. This would save the deceleration half of an interstellar spacecraft's fuel, and provide an auxiliary propulsion system in the target solar system. Interstellar space contains very small amounts of hydrogen. A fast-moving sail would ionize this hydrogen by accelerating the electrons in one direction, and the oppositely-charged protons in the other direction. The energy for the ionization and cyclotron radiation would come from the spacecraft's kinetic energy, slowing the spacecraft.
The cyclotron radiation from the acceleration of particles would be an easily detected howl in radio frequencies.
Gaseous Fission Reactors One limitation on the specific impulse of conventional fission reactions is that if the temperature is to high the reactor core melts. The solution is to create a reactor whose core is gaseous. This can create specific impulses of 20,000 which is good enough for fast interplanetary travel.
Mini-Magnetospheric Plasma Propulsion
A major problem with magnetic sails is that they would have to be superconducting. To circumvent this problem, NASA has attempted to develop a system using a conductive plasma constrained by the magnetic field, a sort of synthetic magnetosphere.
A prototype is being developed at the Univ. of Washington. It has been successfully tested on earth, but not deployed in space.
This mechanism creates an electromagnetic bubble or mini-magnetosphere around the spacecraft, using a cloud of plasma to conduct the current to create the magnetic field. The solar wind is deflected (being made of protons, it has mass) and the reaction accelerates the spacecraft.
The advantage is that no reaction mass is depleted or carried in the craft. This is a form of magnetic sail. Some amount of material is required to replace the ions of the cloud, and power is required to excite and power the cloud.
The solar wind travels at 300-800 km/s.
In the 1954 explosion at Bikini Atoll, a crucial experiment proved that nuclear explosives could be used for propulsion. Two graphite-covered steel spheres were suspended near the bomb. After the explosion, they were found some distance away, proving that engineered structures could survive a nuclear fireball.
Later this information was used to design a spacecraft propulsion system called "Orion" in which nuclear explosives would be dropped through a graphite-covered pusher-plate, and exploded. The pusher plate would be mounted on large shock-absorbers.
A scale model using chemical explosives was constructed from a soft-drink dispenser, and flew a controlled flight for 23 seconds.
Calculations show that this rocket would combine both high thrusts, and a high specific impulse, a rarity in rocket design.
This is also the only known method of performing manned interstellar exploration with current technology. It would be slow, requiring several generations to get to Alpha Centauri (the closest known solar system other than our own), but it would arrive, assuming it had no accidents.
The salt-water rocket
This rocket would be fueled by water, with dissolved salts of plutonium or U-235. These would be stored in tanks that would prevent a critical mass from forming by some combination of geometry or neutron absorption. The rocket would be powered by a nuclear-thermal reaction when the water was injected into a reaction chamber.
Calculations show that this rocket would have both very high thrusts, and a very high specific impulse, a rare combination of traits in the rocket world.
A number of other proposals use power in the form of electricity or heat. Usually these schemes assume either solar-electric power, or an on-board reactor. However, both power sources are heavy. Therefore, one could instead leave the power-source stationary, and power the spacecraft with a beam of microwaves or a laser beam from a fixed installation. This permits the spacecraft to leave its power-source at home, saving significant amounts of mass.
Microwave broadcast power has been practically demonstrated several times. The first time was at Goldstone California, in 1974.
In this system, a reactor would generate such high temperatures that the light from the reactor would provide thrust. Think of a nuclear light-bulb, with a reflector. The big advantage is that nothing material leaves the spacecraft, so only nuclear fuel is depleted. This has the highest-known specific impulse of any rocket that might be made with known technology. The reactor would be constructed of graphite and tungsten.
Significantly beyond current engineering
There is currently no known method of sustaining a fusion reaction, although the theoretical power can be calculated, and the nuclear reaction has been verified in laboratory experiments and nuclear weapons. The advantage over a fission rocket is less radiation, and thus less shielding. Of the technologies mentioned here, fusion is the most likely to be feasible in the medium term, as steady progress is being made towards self-sustaining fusion reactions.
Proposed in 1960 by the physicist Robert W. Bussard, this system is a variant of a fusion rocket. It would use a large scoop to compress interstellar hydrogen and fuse it. To save mass, some people have suggested using a magnetic field for a scoop. Calculations have shown that the drag of a scoop would be more than the energy generated by the fusion reaction. However, the calculations (by Robert Zubrin and an associate) inspired the idea of a magnetic parachute or sail-- when life rains lemons, make lemonade.
There may be practical modifications of this concept. For example, perhaps one could shoot nuggets of fuel in front of a spacecraft from a fixed base, and then the spacecraft would not have to accelerate its own fuel.
In this system, antimatter would be injected into a target of normal matter. The resulting reaction products are mostly charged particles. They would be focussed by a magnetic nozzle, providing thrust. The chief difficulty is obtaining the antimatter. Antihydrogen has been produced using particle accelerators, but this method is too expensive to be practical at this time.
Requires new principles of physics
Another proposal would create large wormholes, big enough to pass people. One would then transport one end of the workhole to a different place, creating a short-cut. This travel systems might be possible with extremely odd materials like neutronium, strange matter, or negative mass, but no one knows how to get these.
Under some conditions, current physics indicates that it might be possible to move backward in time. For insterstellar flight, one could therefore use a conventional, slower-than-light spacecraft, yet still arrive instantaneously. One such machine would require two neutronium cylinders, spinning at an appreciable fraction of the speed of light. Another would consist of a ring-shaped black-hole, spinning at light-speed. Another would use the twin-paradox of relativity to make one end of a wormhole older than the other. One could then go backwards and forwards in time by going through the wormhole. See also Time Travel.
Ordinary matter can grab space, causing gravity but no one knows how to turn this effect into a space-drive. A physicist, Miguel Alcubierre, proposed a method of stretching space in a wave, in such a way that a ship would ride the wave, but there is no way known to induce such a wave, or leave it once started.