Home / Engineering / Nuclear-thermal rockets
Nuclear-thermal rockets
The Gateway shuttle runs 929 MW of reactor heat through liquid methane from Kraken Mare and throws the hot gas out of its nozzle.
See it in the tour: chapter gateway, shot 2.4.
Heat without fire
A chemical rocket carries its energy in its propellant. A nuclear-thermal rocket gets it from a reactor. In the shuttle's core, a neutron splits a uranium-235 nucleus, the fragments heat the fuel, and propellant flowing through channels in the fuel picks up that heat. The hot gas expands through the nozzle. It never burns physics.
During a burn the reactor runs at 929 MW of heat and gives 300 kN of thrust design.
Light propellant is fast propellant
Heat a gas and its molecules move faster. At the same temperature, light molecules move faster than heavy ones. So exhaust speed rises with temperature and falls with molecular mass physics.
Rocket engineers measure this as specific impulse, in seconds: how long one kilogram of propellant can push with a force equal to its own Earth weight. Higher is better.
- Hydrogen, the lightest molecule, gives about 900 s physics.
- Methane, one carbon and four hydrogens, gives about 600 s challenge.
- Methalox, burning the same methane with oxygen, gives about 330 s at sea level and 360 to 380 s in vacuum physics.
NASA puts nuclear-thermal propulsion at twice the propellant efficiency of chemical rockets physics.
Why the shuttle burns methane
Hydrogen wins per kilogram and loses per cubic metre. The Long Passage's 3,935 tonnes of liquid hydrogen fill 58,354 m³. The shuttle's 206.7 tonnes of methane fit in 514 m³: two tanks, 6 m across and 11 m long design. Methane comes out of Kraken Mare and stays liquid at Titan's 94 K without help physics. The shuttle keeps a hydrogen option for runs that need it design.
The price is mass. The design budget is 4.5 km/s, a round trip plus margin. On methane that takes a mass ratio of 2.149: 206.7 tonnes of propellant for a 180-tonne ship and payload design.
The Gateway run
Leaving the Gateway's 750,000 km orbit costs 0.81 km/s. The shuttle reaches Titan at 1.53 km/s relative to the moon and spends 1.12 km/s to capture into a 1,500 km parking orbit, above the bulk of the haze. One way is 1.92 km/s physics. Aerocapture in Titan's air would cut the capture burn design. The trip takes days.
Heritage
The idea is old. Project Rover began at Los Alamos in 1955, and the NERVA program that followed ran from 1961 to 1973. NASA credits nuclear-thermal propulsion with twice the propellant efficiency of chemical rockets, and hydrogen designs of that lineage are put at about 850 to 900 s challenge.
Concordia's zero-g yard built the nuclear-thermal Meridian class between 2041 and 2045: about 900 s and about 4 km/s on full tanks fiction. Hydrogen volume set Meridian 04's 700 m class length. Meridian-class deliveries assembled the Gateway's truss in 2071 fiction. The shuttle is a Meridian descendant, with the same capture-frame hard points and livery logic design.
What you see
Almost nothing leaves the nozzle that the eye can catch. The drama is the white-hot nozzle, reactor glow through the inspection ports and, on hydrogen, a faint pink hydrogen-alpha line at 656 nm physics.
Key numbers
| Quantity | Value | Basis |
|---|---|---|
| Specific impulse, hydrogen | about 900 s | physics |
| Specific impulse, methane | about 600 s | challenge |
| Reactor heat / thrust | 929 MW / 300 kN | design |
| Ship and payload (dry) | 180 t | design |
| Methane load / volume | 206.7 t / 514 m³ | design |
| Design delta-v | 4.5 km/s | design |
| Gateway to Titan orbit, one way | 1.92 km/s | physics |
| Titan parking orbit | 1,500 km | design |
| Rover / NERVA | from 1955 / 1961 to 1973 | physics |
| Shuttle length / diameter | 45 m / 9 m | design |