The interstellar spaceship that is used in the Avatar movie story is based upon the Valkerie Antimatter Starship designed by Charles Pellegrino and Brookhaven physicist Jim Powell
Dr. Pellegrino served as a scientific consultant on James Cameron's Avatar movie. The interstellar vehicles seen in the film are based on the designs of Pellegrino and Powell's Valkyrie rockets, fused with Robert L. Forward's designs.
Project Valkyrie at wikipedia
Instead of a solid spacecraft with a rocket at the back, Valkyrie is built more like a train, with the crew quarters, fuel tanks, radiation shielding, and other vital components being pulled behind the engine on long tethers. This greatly reduces the mass of the ship, because it no longer requires heavy structural members and radiation shielding.
Initially the Valkyrie's engine would work by using small quantities of antimatter to initiate an extremely energetic fusion reaction. A magnetic coil captures the exhaust products of this reaction and it is expelled with an exhaust velocity of 12-20% the speed of light (35,975-58,900 km/s). As the spacecraft approaches 20% the speed of light more and more antimatter is fed into the engines until it switches over to pure matter-antimatter annihilation. It will use this mode to accelerate the remainder of the way to .92 c. Pellegrino estimates that the ship would require 100 tons of matter and antimatter, with an undetermined excess of matter to ensure the antimatter is efficiently burned, although another calculation suggests that to reach a speed of .92 c and decelerate afterward Valkyrie would require a mass ratio of 22 (or 2200 tons of fuel for a 100 ton spacecraft).
The chief feasibility issue of Valkyrie (or for any antimatter-beam drive) lies in its requirement of quantities of antimatter fuel measured in tons. Antimatter cannot be produced at an efficiency of more than 50% (that is to say, to produce one gram of antimatter requires twice as much energy as you would get from annihilating that gram with a gram of matter). Since a half a kilogram of antimatter would yield 9×1016 J if annihilated with an equal amount of matter, this quickly adds up to enormous energy requirements for its production. To produce the 50 tons of antimatter Valkyrie would need would require 1.8×10^22 J. This is the same amount of energy that the entire human race currently uses in about forty years.
This may be solved by creating a truly enormous power plant for the antimatter factory, probably in the form of a vast array for solar panels with a combined area of millions of square kilometers. Alternately the antimatter-fusion hybrid drive the Valkyrie uses to accelerate up to 0.2 c would require much less antimatter and, with an exhaust velocity of 30-60,000 km·s-1, still compares quite favorably with competing engines such as the inertial confinement pulse drive used by Project Daedalus or Project Orion (nuclear propulsion).
Charles Pellegrino's website has more information
The engine is simply a magnetic coil, which generates a magnetic field, against which particles from the matter-antimatter reaction zone are bounced. The magnetic field (and hence the coil), is propelled forward by the bounce. The coil then pulls the rest of the ship along on a string, much as a motorboat pulls a water skier. A pulling rather than a pushing engine eliminates most of the structural girders that would not only, by their mere existence, add unwarranted mass, but would multiply that mass many times over by their need for shields and cooling equipment, and by added fuel to push the added fuel . . . leading to a chain reaction of design complications. . .and to an engine that burns hotter, but which cannot afford to push the giant to even a significant fraction of lightspeed. By contrast to what has traditionally become known as the large, slow-moving 'space ark' approach to interstellar flight, Valkyrie becomes a low mass speedboat
Valkyrie is the ultralight of rockets, consisting mostly of naked magnetic coils and pods held together by tethers. Indeed, it can best be summed up as a kite (with magnetic field lines instead of paper sheets) that flies through space on a muon wind of its own creation. Valkyrie's fuel stores (both matter and antimatter combined) are estimated at slightly less than half the mass of the rest of the spacecraft, or about one hundred tons. technology. The Valkyrie will have a maximum cruising speed of 92 percent lightspeed.
A suggested approach is made for the generation of the antimatter, but there are several ways advanced technology could be used to produce a lot of antimatter. The Valkyrie has some similarities to the Vasmir.
Mr. Pellegrino's response to Adam Crowl:
On Valkyrie, the lower mass of material you were quoting was for up to 10%c - much lower than the mass for giants like Daedalus, and other such nonsense. The mass of propellant is kept low because up to about 10% c you can go with the lower exhaust velocities of antiproton-triggered fusion.
In any case, the antiproton triggered fusion system, scaled down to Valkyrie Mark II, is wonderfully practical for getting around the solar system at a mere 750 km/sec. (this velocity would eventually be practical for Project Spaceguard: the kinetic force of merely ten Toyota masses impacting a comet or asteroid at this velocity (diameter 1/4 mile) would completely "dust" the object.
In answer to original question, for a true, Valkyrie Mark III (requiring direct proton-antiproton annihilation after 15%c), interstellar crewed mission, the propellant mass would of course exceed the ship mass. After 92%c, the excess becomes too extreme.
Mark III Engine
Mark III is quite simple, actually. Temperature regulation in the antimatter pods will control how fast or slow antihydrogen white flake is permitted to evaporate. As the evaporated antihydrogen leaves the magnetic bottle, and is guided toward the magnetic gun barrel of an atomic accelerator, the atoms are ionized and stripped of their positrons. The positrons are simply ejected into space (for, if allowed to react with electrons, they will produce powerful gamma rays while providing essentially zero thrust). The antiprotons are accelerated to approximately 750 kilometers per second, and when they arrive at the reaction zone, behave somewhat like slow relativistic bombs (mark this as an oxymoron, albeit an essential one). At this velocity, the antiprotons pass like ghosts through beryllium windows, hardly noticing that they have passed through anything at all. They detonate when they reach ('and stick to') the hydrogen nuclei behind the window (including deuterium, and possibly traces of tritium), and by carefully controlling the number of antiprotons reaching the hydrogen (by regulating evaporation rates in the antimatter pods), and hence controlling the temperature of the hydrogen target, the result becomes a finely tuned fusion reaction – in effect, an antimatter triggered hydrogen bomb that, instead of exploding, merely glows, at any rate one wants it to glow.
That glow is in fact a spray of (for our purposes) reasonably massive charged particles, among them helium nuclei. Just as the antiprotons shooting in through the beryllium window fail to notice that a container wall exists, any fusion products shooting out (at the still relatively slow velocity of 12 to 20 percent lightspeed), depart like beams of light exiting glass. The particles then bounce off the ship's forward magnetic field, giving away their energy as thrust.
As the ship's speedometer begins to climb above twelve to twenty percent the velocity of light, fusion ions, though more massive than the products of straightforward proton-antiproton annihilation, decline significantly in propulsion efficiency. To push the Valkyrie to a higher fraction of lightspeed, higher exhaust velocities are needed. At this point, the Mark III reaction mix depends less and less upon fusion, until ultimately it shifts purely to proton-antiproton pairing. At this point, the less efficient reaction (which sheds low mass particles at high speed), has become the more efficient reaction, if for no other reason than it is our only choice.
The reaction products, traveling at high relativistic speed (the speed we want to get our rockets up to) consist of elementary particles called mesons. Each meson has a mass intermediate between a proton and an electron. It is essentially a proton fragment gone so relativistic (read ballistic) that it is at once a particle and a wave, and some of its quarks and gluons have dispersed into the universe as energy (read, massless photons and neutrinos). The matter-antimatter spray produces three varieties, or 'flavors,' of pi-mesons.
1. Neutral pi-mesons comprise thirty percent of the proton-antiproton reaction products. They decay immediately into gamma rays.
2. Positively charged pi-mesons, traveling near the speed of light, decay into positively charged mu-mesons (muons) and neutrinos after flying, on average, only twenty-one meters. The muons last several microseconds (almost two kilometers) before decaying into positively charged electrons and neutrinos.
3. Negatively charged pi-mesons behave the same way positively charged ones do, except that the resulting muons and electrons are negatively charged.
The charged pions and muons are the particles we want, and preferably we want the innermost fringes of the engine's magnetic field (or magnetic pusher plate) to reach within twenty-one meters of the reaction zone, so that it can steal whatever thrust the pions have to contribute before a significant fraction of them have decayed and shed some of their energy as useless neutrinos.
The BBC also has a page with information on the Valkyrie
Avatar Movie Spaceship Backstory
Here is the link to the ISV Venture Star page on the Pandorapedia
Antimatter Tutorial Video
NASA Antimatter Production Paper
28 page pdf - Antimatter Production for Near-term Propulsion Applications G.R. Schmidt, H.P.Gerrish and J.J. Martin
The development of antiproton Penning traps has progressed extremely well over the last 10 years. The PS200 experiment trapped over 10^6 antiprotons for periods of hours. This is seen as a means of soon being able to confine up to 10^12 antiprotons with transfer to a remote site for periods of many days. Synergistic Technologies of Los Alamos, NM is currently developing a magnetic degrading spectrometer which will simply and inexpensively decelerate antiprotons into such portable traps. In this case, a more efficient decelerator section will be required to achieve production rates equivalent to ~1 microgram per year. Antiproton decelerators which accomplish this do exist (e.g., at CERN), and in the case of FNAL would cost about $10 million to construct.
1980's a RAND Corporation study of antimatter production - antimatter production capacity of 0.1 to 1 gram per year could be achieved with a new machine costing $3 to $10 billion.