|Fig. 1: The NERVA thermonuclear rocket developed by NASA in the 1960s. (Source: Wikimedia Commons)|
Nuclear propulsion and power have been attractive but mostly unrealized options for spaceflight that have been under consideration roughly as long as space programs have existed. For space vehicles of all types, mass is a major concern given how difficult and expensive it is to transport that mass into space, so fuel and power sources should be as energy-dense as possible. Additionally, for missions beyond Earth orbit, the distances to be traveled are so great that long-lasting, high thrust and becomes important if the vehicle is to reach its destination in a reasonable amount of time, especially for a manned mission (high mass and power requirements, plus longer exposure to cosmic rays if the vehicle moves slower).  The high energy density of nuclear fuel makes it an obvious candidate for this application, although the methods of employing it are at varying levels of technological readiness.
The fission of uranium or plutonium atoms releases large amounts of energy, so it can be used to heat propellants such as hydrogen so that they expand out through a nozzle at high speed, generating thrust.  Hydrogen's low atomic mass allows it to accelerate quickly out the nozzle and provide more thrust than larger elements would. Fission reactors are relatively well understood and used commercially for electrical power, but the rocket would replace the power plant's water with hydrogen.  The United States has been considering this type of rocket for over half a century now, and has produced not only designs but also ground tests and studies that demonstrate the technology, but this work only lasted from 1955 until 1973, when funding and interest in the relevant missions dried up.  These rockets were designed with manned missions to Mars (and possibly other planets) in mind, since a trip to Mars could take only 6 months with a fission rocket (depending on the timing of the launch).  The journey would take roughly half the travel time of a chemical rocket, reducing astronaut exposure to cosmic rays, assuming the radiation from the reactor itself can be sufficiently shielded. Fission-based nuclear thermal rockets are the low hanging fruit when it comes to nuclear propulsion, making them a top candidate for the first manned missions to other planets.
|Fig. 2: Project Orion riding the shock wave of a nuclear explosion. (Source: Wikimedia Commons)|
While traditional rockets keep the nuclear reaction inside the vehicle, nuclear pulse fission uses a series of nuclear explosions outside the spacecraft to propel the vehicle forward on the shock waves. The spacecraft would have to carry a collection of small nuclear bombs to release behind itself every few seconds to achieve the desired changes in velocity. [2,3] Each bomb would use its explosion to accelerate some of its mass toward the rear of the spacecraft, where the products of the explosion would impact a pusher plate and transfer the force to the spacecraft. [2,3] Project Orion, which lasted into the 1960s, is credited with this idea, which it developed on paper but never built before the project was cancelled due to nuclear test ban treaties, reflecting concern about nuclear fallout from the detonations. [2,3] The design nonetheless presented a relatively cheap method of reaching the outer planets in short enough mission times to accommodate astronauts, all without requiring significant advances in technology.  A later version of Orion even aimed to reach other stars using fusion explosions, and it was calculated that the vehicle could be propelled to velocities of up to 10,000 km/s (~3% the speed of light).  This design is based on existing technology and would be doable if humanity felt pressed to it (if, for example, the planet needed to be evacuated), in which case a thermonuclear bomb-propelled ark could take us in search of a habitable planet around another star.  Short of that, however, the fear of nuclear fallout due to near-Earth detonations and testing may postpone this design indefinitely.
When light atoms like hydrogen fuse together, the energy released surpasses that released by splitting heavier atoms through fission, so much so that a rocket propelled by this energy could reduce the travel time to Mars from a matter of months to a matter of weeks.  However, fusion is not as well understood or as easily controlled as fission, which is why fusion reactors aren't used in commercial power plants, but only for research. Fusion reactors still struggle to contain their plasma within magnetic fields, and current designs require so much energy to sustain the fusion reaction that the reactors produce a net loss of energy. Nonetheless, fusion releases so much energy that the technology only needs to get a tiny bit past the break-even point before it can generate considerable thrust or electrical power. 
|Fig. 3: Project Daedalus with its colossal fuel tanks. From nose to nozzle, the design was nearly twice the length of a Saturn V rocket. (Source: Wikimedia Commons)|
Fusion rockets differ from fission rockets in that the propellant expelled through the nozzle would be the hydrogen-helium plasma involved in the fusion itself. Like in a tokamak fusion reactor, the plasma would need to be contained by a magnetic field, except the field would be shaped to allow the plasma to "leak" out the nozzle.  However, in order for the particles generated by the fusion reaction to be charged so that the magnetic field can direct them out the nozzle, the plasma must be heated much more than is necessary simply to initiate fusion. About 100 million Kelvin is the required temperature for fusion, but the reaction mainly produces neutrons at this temperature, which can't be steered in a certain direction for thrust. 600 million Kelvin, on the other hand, can enable a different fusion reaction that produces charged alpha particles, meeting the requirement for thrust. The plasma can be heated to this point using microwaves, although such heating would require significant power on board the spacecraft. 
An alternative to the continuous-flow rocket described above is Project Daedalus's pulse propulsion engine, where pellets of fusion fuel would be injected and then ignited by electron beams inside a reaction chamber. The fusion fuel included helium-3 for its high energy release, although this isotope is extremely rare on Earth and would probably need to be mined from the moon or the atmosphere of gas giants. [3,5] Daedalus was designed in the 1970s to be a fusion-propelled interstellar mission that would reach a max speed of 12.2% the speed of light and arrive at Bernard's Star (5.9 lightyears away) after 50 years, all using nothing too far beyond the existing 1970s technology.  Although the project was simply a feasibility analysis to be published and not a physical vehicle to be tested, a reasonably sound (though expensive) design does exist, and other designs and studies have followed it.  Fusion is arguably the only propulsion technology within reach of our current knowledge of physics that is capable of an interstellar mission lasting less than an average human lifetime, even though significant advances in fusion reactor technology are still a prerequisite.
|Fig. 4: The Bussard interstellar ramjet, scooping hydrogen from its environment. (Source: Wikimedia Commons)|
Instead of carrying all the fuel for a mission inside the vehicle, the nuclear ramjet (also known as the Bussard Ramjet) would use an electromagnetic "scoop" to collect hydrogen from space for use in nuclear fusion.  The shape of the flow path constricts (similar to an air-breathing ramjet), causing the gas to compress until fusion begins. Since the density of hydrogen in space is projected to be only at most 1-2 atoms per cm2 on any given plane through space, the scoop would need a large cross-sectional area: for a circular scoop, a radius on the order of 60 km may be necessary to maintain the necessary thrust.  Additionally, the thrust produced must be high enough to overcome the drag force caused by scooping particles up from the environment. However, if thrust does overcome drag, the ramjet has the potential to travel incredibly fast, with up to 1g acceleration that could in theory bring the velocity close to the speed of light, since the spacecraft may be able to avoid ever running out of fuel.  At such speeds, time dilation would make the journey seem noticeably shorter to anyone on board than it would to someone back on Earth, to the point where the hundreds of thousands of years it would take to reach the center of the galaxy could be traversed within the crew's lifetime.  This technology would make sense for a one-way (possibly colonizing) mission to a far-off habitable world. The physics behind this design is reasonable, but the technology to generate such an electromagnetic scoop has yet to be developed, not to mention the necessary advances in fusion reactor technology.
Overall, there are promising long-lasting, high-energy, high-thrust nuclear technologies that have been studied for space applications over the years, but due to obstacles such as funding cuts and public concern about the safety of nuclear technology, many of them were never fully developed. Nonetheless, much of the knowledge and technological ability needed to begin implementing these systems remain, should humanity decide that that the risk is not too great and that space exploration (or space travel for any reason) is worth the expense.
© Ashley Micks. The author grants permission to copy, distribute and display this work in unaltered form, with attribution to the author, for noncommercial purposes only. All other rights, including commercial rights, are reserved to the author.
 S. K. Borowski et al., "Nuclear Thermal Rocket/Vehicle Design Options for Future NASA Missions to the Moon and Mars," U. S. National Aeronautics and Space Administration, NASA-TM-0107071, Septemper 1993.
 W. E. Moeckel, "Propulsion Systems for Manned Exploration of the Solar System," U. S. National Aeronautics and Space Administration, NASA TM X-1864, August 1969.
 G. R. Schmidt, J. A. Bonometti and P. J. Morton, "Nuclear Pulse Propulsion: Orion and Beyond," Am. Inst. Aero. Astro., AIAA 2000-3856, July 2000.
 D. Graham-Rowe, "Nuclear Fusion Could Power NASA Spacecraft," New Scientist, 23 Jan 03.
 C. H. Williams et al., "Realizing '2001: A Space Odyssey': Piloted Spherical Torus Nuclear Fusion Propulsion," U.S. National Aeronautics and Space Administration, NASA-TM-2005-213559, March 2005.
 R. W. Bussard, "Galactic Matter and Interstellar Travel," Astronautica Acta 6, 179 (1960).