Could a nuclear thermal rocket take off from the surface of the Moon?

The application is a shuttle from LEO to the Lunar surface.

Fairly recent NASA plans include an NTR LEO to LLO shuttle, but use a separate lander to descend to the Moon's surface. https://ntrs.nasa.gov/archive/nasa/casi.ntrs.nasa.gov/20170005398.pdf

Dan Hanson's suggested link on the Lunar Ice NTR is much appreciated. This rocket is optimized for lunar ascent. It is desired to consider the inverse problem.

Could an NTR shuttle from LEO to Lunar orbit be expected to land and ascend from the Moon? If it disappeared around to Farside, and did not reappear, could it be assumed to have landed, or would knowledgeable observers think it unlikely that such a shuttle would descend?

  • 5
    $\begingroup$ Is its thrust to weight ratio > 1? $\endgroup$ Feb 19, 2020 at 18:37
  • 1
    $\begingroup$ + is the mass fraction sufficient $\endgroup$
    – user20636
    Feb 19, 2020 at 20:06
  • 3
    $\begingroup$ This was closed while I was answering, so I'll post a short answer here: There's no real problem with that, conceptually. In fact, there is already a preliminary design study for a nuclear thermal lunar ferry that uses steam exhaust from lunar water superheated by a nuclear thermal motor to fly from the lunar surface to lunar orbit and back. According to the paper, a nuclear thermal steam rocket could deliver 60 tons of payload to lunar orbit, and require about 252 tons of water propellant from melted lunar ice. LEO would require more water mass. neofuel.com/moonicerocket $\endgroup$
    – Dan Hanson
    Feb 19, 2020 at 22:21
  • 2
    $\begingroup$ Thank you so much, Dan Hanson! That's pretty near exactly what I was looking for. The reference has much useful discussion. This spacecraft may appear in my new novel. $\endgroup$
    – DrBunny
    Feb 19, 2020 at 22:39
  • 2
    $\begingroup$ @DanHanson It's always frustrating when insta-closers prevent others from posting a good answer. When there are 4 close votes and I want to answer, I begin writing a bit and then quickly post it with a short explanation that I'm currently editing. You can continue to edit answer posts after the question is closed. $\endgroup$
    – uhoh
    Feb 20, 2020 at 2:05

2 Answers 2


I originally voted to close the question as I thought it was a bit broad (Can we build an NTR that can lift off from the Moon? Sure!), and because of the response to Organic Marble's and JCRM's comments (which were hints on how to improve the question).

But, since it's been re-opened...

Given the link to Sorensen's blog post, we'll assume we're talking about the same kind of SSTO NTR he is, and as a starting point I'll use his numbers for specific impulse, propellant-mass-sensitive term, and gross-mass-sensitive term. I just replaced the deltaV for lunar escape velocity, and get a payload mass fraction of 0.3125.

So yeah, you can build such a system to lift off from the Moon (with some useful payload) and head to Earth.

So, the next obvious questions would be:

  • What kind of NTR? Solid core? Gas core? Something else?
  • What's the source of the propellant? Ice from the lunar poles? Or something else?
  • What's the source of the fissionable material?
  • What are you shuttling between Earth and the Moon? People? Stuff? How much?
  • How many trips do you expect to get out of a given reactor fuel load?
  • What sort of environmental constraints would we be under? It's not like the lunar surface is a friendly environment to begin with, but would contamination from the rocket exhaust be an issue for other surface operations?
  • $\begingroup$ Should I edit the question, or add a comment? I don't want to consider fuel sources or cargo, just whether an NTR shuttle from LEO to Lunar orbit could be expected to land and ascend from the Moon. If it disappeared around to Farside, and did not reappear, could it be assumed to have landed, or would knowledgeable observers think it unlikely that such a shuttle would descend? The ice rocket is heavily optimized for Lunar ascent, but the discussion and link to it are much appreciated. I don't have a source for an LEO-Lunar orbit NTR shuttle design to approach the question from that viewpoint. $\endgroup$
    – DrBunny
    Feb 21, 2020 at 23:19
  • 1
    $\begingroup$ The issue with most NTR designs I’m aware of is thrust. Environmental issues aside, NTRs have poor thrust to weight ratios - reactors are heavy, and hydrogen’s low density doesn’t help. Again, going with Sorensen’s numbers and plugging in lunar gravity, such a shuttle could land on the Moon and ascend (after taking on more propellant). It would not be able to lift off from the Earth’s surface, though, and I’m not sure it could soft-land. Now, there may be NTR designs I’m not aware of with better TWR, which could work both places. $\endgroup$
    – John Bode
    Feb 22, 2020 at 2:32
  • $\begingroup$ The LOX-Augmented NTR, or LANTR in the Borowski paper "By varying the oxygen~to-fuel mixture ratio (MR), the LANTR concept can provide variable thrust and specific impulse (Isp) capability with a LH2-cooled NTR operating at relatively constant power output. For example, at a MR = 3, the thrust per engine can be increased by a factor of 2.75 while the Isp decreases by only 30%. $\endgroup$
    – DrBunny
    Feb 22, 2020 at 3:30

The Scorpion "General Purpose Space Transportation System" design, with an estimated program cost of $160bn would be a six person vehicle, 107 m long with a dry mass of 230 tonnes capable of lunar landing carrying six 20 tonne cargo pods and an 11 tonne Anzu capsule (to be developed as part of the costs) used as an escape pod

It would be powered by a 40 tonne Serpent-H Nuclear thermal electric engine design, with an expected thrust of around 2000 kN, and 12746 seconds specific impulse.

Depending on the cargo carried, it could probably take off with that engine (given some sort of launch rails), however, mission constraints forbid use of the engine under 50km altitude above the lunar surface to ensure the reactor is not active within the potential range of danger to the inhabitants of surface bases. So it would use hydrolox engines with combined 2400 kN of thrust, which are required for landing due to lack of throttling and minimum firing duration requirements.

(From JBIS vol 72 number 7)

It is designed to be launched by the (also non existent) Skylon reusable SSTO HOTOL spaceplane.

  • $\begingroup$ You might consider making it more explicit that the vehicle and engine don't exist. $\endgroup$ Feb 24, 2020 at 3:00
  • 1
    $\begingroup$ do you think calling it "a British design" is explicit enough that it'll never exist @RussellBorogove $\endgroup$
    – user20636
    Feb 24, 2020 at 7:52
  • $\begingroup$ @JCRM you made me laugh out out in a library! grrr..... ;-) $\endgroup$
    – uhoh
    Feb 24, 2020 at 10:27
  • $\begingroup$ The question is for a nuclear thermal rocket, not nuclear electric. $\endgroup$
    – DrBunny
    Feb 25, 2020 at 5:21
  • $\begingroup$ This is a nuclear thermal electric @DrBunny - instead of being LOX augmented as LANTR the hot hydrogen is then further heated as in an arcjet $\endgroup$
    – user20636
    Feb 25, 2020 at 10:02

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.