On @ The 90: NASA Re-Reaches for the Nuclear Future

NASA is revisiting sources of propulsion that could greatly aid crews visiting distant locations in our solar system. Photo Credit: MSFC / Emmett Given
NASA is revisiting sources of propulsion that could greatly aid crews visiting distant locations in our solar system. Photo Credit: MSFC / Emmett Given

Research at the Marshall Space Flight Center has picked up on an old thread to advance rocket technology. Known as the Nuclear Thermal Rocket, this technology uses the heat generated by nuclear reactions to produce rocket thrust. As the space agency once again eyes destinations beyond the orbit of Earth, NASA must review sources of propulsion better suited for long-duration missions to distant destinations—and that means nuclear.

Normal chemical rockets use the heat generated by chemical reactions between fuels and oxidizers to create hot gas which is forced through a nozzle to produce thrust. Nuclear thermal rockets use the heat generated either by the radioactive decay of unstable isotopes or nuclear fission reactions to heat gas, force it through a nozzle, and create thrust. The key difference is that nuclear thermal rockets are more efficient at harvesting the energy of their heat source, and thus can produce higher thrust and greater exhaust gas velocities—two key measures of rocket performance. The Space Shuttle Main Engine’s combustion temperature is about 3500 K. While NASA-Marshall’s simulated NTR has a similar internal temperature, real NTRs can get much hotter, as hot as the materials that make them up can stand.

An image of an Space Shuttle Main Engine being test fired at Stennis Space Center. Image Credit: NASA / Stennis Space Center

NASA-Marshall’s NTREES (Nuclear Thermal Rocket Element Environment Simulator) experiments do not involve actual radiological materials, but simulated radioactive substances. The engineers use materials that might be used to contain the reactor-fuel and heat it to the incredible temperatures an NTR rocket would endure. They also pass cryogenically cold hydrogen over the containment element to simulate the way the rocket would produce thrust. NTRs are not a new idea by any means. Science fiction writers speculated about nuclear engines during the 1940s and ’50s. NASA itself, in concert with several other federal agencies, produced a working version of an NTR under the NERVA (Nuclear Engine for Rocket Vehicle Application) project. The final product, the NRX/XE engine, was considered safe enough to use on an actual manned Mars mission, until Congress and then-President Richard Nixon canceled the program by slashing the budget for manned Mars exploration.

But the Nuclear Thermal Rocket is the key to the future. Consider now the expense of using expendable rockets over and over—not to mention the accumulation of space junk as spent upper stages hang around in useless orbits over the Earth, sometimes breaking up or even exploding spontaneously and creating more debris, which can go on to threaten more spacecraft and create more debris.

Now imagine a reusable upper stage powered by an NTR. The “space-tug,” as the concept has long been known, and an NTR-powered one would only need one propellant. Liquid hydrogen would be best (in general, the lower the atomic weight of a propellant, the better it is for use in rocketry), but methane could work as well.

In such a scenario, a small expendable carrier rocket would launch a payload into LEO, where the space-tug would rendezvous with it and carry it up to the correct orbit, then maneuver away to intercept its next payload.

Such a tug would need only the occasional refueling, as its engine might last over a decade.

It could also be of enormous use to the manned spaceflight program. NERVA meant for its engines to send men to Mars, and Marshall’s NTREES experiments have the same ultimate purpose, but such an engine could meet needs closer to home as well.

The problem with the Apollo program was that every time we needed to put a man on the Moon, we had to launch a massive Saturn V rocket. What if, instead, we could launch only a medium-lift rocket and send a crewed capsule to a space station at the Earth-Moon L1 point (the point in space where the Earth’s and the Moon’s gravity balance out, and an object placed there would tend to stay there)? The capsule could rendezvous with the station, which might or might not be permanently inhabited. Think of the science that could be done at such a station—Earth observation, Moon observation, stellar astronomy, solar astronomy, solar wind physics … the possibilities are endless—which would have one or more reusable lunar landers docked, each using these NTR engines. The space station would double as a fuel depot for them. From that station, any point on the Moon could be explored at any time with no need for a Saturn V/SLS class launch from the ground. In fact, teams could be maintained on the Moon as on the ISS, switching out via the reusable landers and ascenders and the capsules to carry them back to Earth, or an Earth-orbital hub station.

And, of course, there is the engine’s usefulness in a manned Mars mission. By allowing a near-constant acceleration (there would be need for only one propellant to be carried, instead of the greater weight, complexity, and expense of multiple tanks to maintain multiple propellants) and the greater thrust of the engine, the travel time to Mars could be slashed enormously.

Artist's representation of an NTR-based Mars exploration ship
An artist’s depiction of how a manned NTR-based Mars exploration vehicle might look. Image Credit: NASA/Marshall Space Flight Center

And even unmanned exploration could benefit from the boon of an NTR. Right now, the greatest hurdle in the way of exploring the outer solar system is the time it takes probes to get to their destinations. In that time, the teams that built and operated the craft grow old, retire, and even die, and the spacecraft components themselves are exposed to degrading radiation and the harsh environment of space. By the time they arrive at their targets, they have already lost much of their useful lives.

NERVA nuclear/LH2 engine
This is an image of the completed NERVA test stage. Image Credit: NASA Found at: Encyclopedia Astronautica www.astronautix.com

Consider the New Horizons probe, launched in 2006 and not expected to arrive at Pluto until 2015—and this is the fastest object humans have ever launched into space. Furthermore, it won’t even be able to stop at Pluto and carry out a thorough characterization of the planetary system. It’s traveling too fast and has too little fuel to slow down, so the best we can hope for are some fleeting glimpses as the spacecraft tears by the dwarf planet at a relative velocity of 13.8 km/s.

And Uranus and Neptune have yet to have dedicated probes visit them, and, again, time is the greatest hurdle (not to dismiss the budgetary hurdles, however). But a Uranus or Neptune probe with an NTR engine could arrive in a reasonable amount of time. At orbit insertion, the NTR stage could be ditched into the target planet (like Galileo’s RTG into Jupiter) or kept and used to make the kind of drastic orbital maneuvers Galileo and Cassini could never make. An NTR-powered Uranus or Neptune probe could sweep close enough to the cloud-tops to get an actual sniff, while still having the power to explore the planets’ moons in detail.

The thrust available to an NTR stage would allow a heavier payload to be sent to the target. Perhaps that would enable smaller “carry-on” probes, like the Galileo Jupiter entry probe and the Huygens Titan landing probe. Such probes could enter the atmospheres of their targets or land on their moons. A sufficiently resilient probe might even be able to endure direct exploration of the outer planets’ ring systems. An NTR-powered spacecraft bus might enable a whole fleet of probes to explore a single planetary target.

But right now all of this remains in the future. The biggest problem is that Americans are inherently afraid of the words “radiation” and “nuclear.” A systematic educational campaign could ease baseless fears about NTR-driven space-tugs, manned-mission engines, and planetary-probe engines, and a write-in campaign to Congress urging them to support the future of humanity itself by supporting the future of space exploration could be vital.




  1. NTRs are somewhat less than the holy grail of engines. Contrary to the article, exhaust temperature must remain considerably lower than the hardware, which in turn must remain considerably cooler than the material thermal limits. Chemical rockets actually run hotter than NTRS. With regenerative and some film cooling, chemical rockets normally run considerably hotter than the materials could stand without active cooling.

    Where NTRs shine is in the ability to use a very low molecular mass exhaust with hydrogens’ molecular mass of two compared to an SSME exhaust in the ten or so range. If the SSME ran stochiometric, then its’ exhaust would have an average molecular mass of eighteen. Isp, which is a measure of exhaust velocity, goes as the square root of temperature over molecular mass. The SSME molecular mass in the ten range is five times that of an NTR.

    If they ran at the same temperature, the Isp of the NTR would be roughly the square root of ten times the Isp of the SSME. Very roughly Isp of 450 for the SSME become Isp of 1,400 for the NTR, or a 3.1 ratio. Since the NTR cannot reach those temperatures, the real ratio becomes 2-2.5. While this is a major improvement, it does not make the solar system into a quick trip to anywhere.

    One gravy feature of some NTRs is the ability to use any reaction mass from methane to carbon dioxide to plain water from anywhere in the solar system. Isp drops below chemical rockets with the heavier chemicals though.

    • I beg your pardon, but on what basis do you insist that an NTR cannot get as hot as the SSME. NASA-Marshall’s NTREES experiment was running at something like 3500 K, right up there next to the SSME.
      An NTR can get as hot as wants just so long as the fuel material or containment materials don’t melt.
      And the fact that it can run with only one propellant chemical saves a lot of mass given tankage and plumbing requirements.
      And you can use it as a power source by diverting some of the hot gas through a turbine. You could even go “staged combustion cycle style” and return the hot gas to the thrust chamber. Or you could just use thermocouples, depending on what kind of current you need.

  2. On the basis of the published material on the subject.

    Tankage and plumbing on pure hydrogen is non-trivial compared to chemical rockets. LOX/LH2 tankage will about a third of the volume of the NTR tanks for the same mass. LOX plumbing and pumping is trivial compared to hydrogen handling.

    NTRs are also notorious for having a low thrust/weight ratio due to all the hardware to make them work. I didn’t say that there was no use for NTR. I said that the article overstated the case.

    Multiple technologies will open the frontier. Eliminating the expended upper stages as you suggest would be a good start with your NTR tug doing yeoman service. It would have to be refueled with hydrogen on a regular basis though, even if it didn’t need nuclear material very often.

    • But why did you say an NTR can’t get as hot as an SSME?

      As to tankage, one LH2 tank and associated plumbing on an NTR is probably cheaper, equally as expensive, or only slightly more expensive than the tankage on a chemical rocket for the same delta-v. The NTR can have a higher specific impulse and greater thrust than a chemical rocket, and you don’t have to fuel it as often.

  3. An NTR can’t get as hot as a chemical rocket because the chemical rocket can use active cooling to keep the combustion temperatures from heating the walls very much. Several companies build chemical rocket chambers from aluminum with active cooling. That option is denied NTRs because the heat must transfer from the walls to the fuel, which means the walls must be as hot as the materials can stand without degrading strength.

    Higher Isp is a given. Higher thrust is not. NTRs have a lot of mass in heating elements and exchangers that don’t exist in chemical rockets.

    As far as tankage cost and mass per Delta-V, We would have to do more trade study on the specific missions to find out if we agree or not that either of us probably are motivated for. It should at least be agreed that an NTR engine will have more cost per engine thrust than a chemical unit.

    It has been a few years and a move since I compared the two, and my references on it are in storage. I have learned to not trust web information beyond my ability to verify if it is important.

    I once posted a chemical/tether/nuclear engine idea to get on the order of 850 seconds with H2/O2 with assist while retaining the NTR ability to use any available volatiles for fuel with Isp higher than chemical. Note the give and take in comments. Conceptual ideas are fun,but not always practical.


    • But an NTR has active cooling. The cryogenically cold hydrogen/whatever you want is coming in, taking away engine heat, and dumping it overboard while providing thrust. Yes the materials need greater heat tolerance, that does not necessarily mean heavier materials. Titanium can stand up to more heat than steel while still being much lighter.

  4. Wow it sounds like we need someone like Robert Bigelow to “license” the NTR tech from NASA put a couple of million into it and make a viable system to go along with the TRANSHAB tech he bought for $2 Million and is now selling back to NASA for the ISS…Then again maybe Elon Musk would be the better choice…either of these two options would be better at maturing this technology than NASA….Wouldn’t it?

  5. No. Let MSFC and Oak ridge do this because they have experience. Let alt.space service ISS and launch comsats for now.

  6. Does a low-thrust engine more than, say, a thousand km out, need LIQUID hydrogen? A gas-bag would need to be clever not to leak, but a gaseous – warm – system still might be lighter and simpler than keeping hydrogen liquid long-term.

    Any thoughts, please?

Atlas V Prepares to Loft NASA’s TDRS-K Satellite into Orbit (Part 2)

Remembering Apollo 1