Maybe not but it's pretty significant. From 350 to 900 takes the payload mass fraction to LEO from 6% to 34%, so SSTO would be practical (though I wouldn't expect to see this for launch). Or, with a 10% mass fraction, the delta-v goes from 7.9 km/s to 20 km/s.
Yes. The difference is definitely a sea change. To put it in concrete terms, a nuclear-powered rocket with sufficient thrust could be single-stage to mars and back for the same propellant budget and mass fraction of a first stage chemical engine which doesn't even get to orbit.
A nuclear thermal engine is not remotely a drop-in replacement for a chemical engine. There are substantial other changes required to the rest of the rocket that add lots of mass and expense, that work to substantially negate a lot of the benefits.
That's true from a political perspective but it's definitely not true from an engineering perspective. Got enough energy to hoist almost anything you like into orbit if you choose your thrust and scale appropriately. As to what happens to the ground underneath, that's a different problem :)
No, extra mass is needed to make them operational, not politically acceptable.
First is engine mass. (All numbers from memory)
A SpaceX Raptor weighs around 4,000 lbs, and generates over 400,000 lbs of thrust (100x). NERVA weighed around 40,000 lbs and only generated 50,000 lbs of thrust (1.2x).
A fully fueled orbital Starship in would weigh around 2.5M with about 25,000 lbs of engine, 150,000 lbs of ship and 2.3M lbs of fuel generating 1.2M lbs of thrust out of three vacuum Raptors.
For a similar level of thrust from NERVA would add nearly 1M lbs of dry mass. Newer designs can cut that engine mass, but at most in half. And your nuclear ship could get by with less thrust and just burn the engine longer. But that brings two new questions
Where does all the heat go?
NTR send almost all of the heat out the back, but not all. Running that engine for hours on end in the vacuum of space (Instead of the very conductive earth NERVA was tested on) is going to heat stress your engine and everything it’s connected to. Do you need to add massive radiators and insulation to run the engines longer?
How do you take off on Mars?
If you have a lot less thrust, you will only be able to lift a lot less fuel and cargo. And you can’t land or take off on Earth. Unlike Starship NTRs may only be usable in between planets, requiring carrying heavy and expensive landers.
Then there is propellant. A 900 ISP is only achievable with hydrogen (H2), using Methane or water or any other feasible propellant drops ISP below 600, making overall system performance far worse than burning methane in Starship Raptors.
OK, let’s use Hydrogen then. Well unfortunately Hydrogen is a shitty propellent in every other measure. It doesn’t compress well so yet requires larger, heavier tanks. Also H2 is a tiny slippery molecule, forcing you to freeze it near absolute zero in order to keep it from leaking, which makes those tanks even heavier. And guess what? It still leaks. Will you have enough fuel left to fire a retro burn to match orbits with Mars? Better leave with at least 30% more than you need.
Now, you probably don’t want to be killed by your engine on the way to Mars, so let’s add some radiation shielding between the crew and the NTR engines, or put the on a long spar sticking out behind the ship. Both add even more mass.
A NTR ship will likely require at least 2-3x more dead mass than a chemical rocket ship, so the question becomes, what’s the point? Why jump through political hoops when Starship VP can already fly crew to Mars in only 90 days and deposit over hundred tons of cargo to the surface of Mars, and cheaply?
A lot of your math is way off, brother. Sure, NERVA was close to 1.0 thrust to weight ratio, but that says a lot more about the state of technology at the time than about the capabilities of nuclear thermal rockets.
The Dumbo / STNP engines were order of magnitude better than NERVA - way more than "half the mass" and I'm pretty sure we can do better than that these days. The quote is "Better than a conventional engine for any given amount of fuel", implying that even with balance-of-system mass issues they're still ahead of chemical rockets using virtually any fuel.
Also, I don't know whether anybody is suggesting crewed missions, or at least I'm not. But "Why bother", well, cost and efficiency and size, frankly. Starship is a bold move, but to me, it looks pretty small. I'd like them to add a zero onto the end of the mass budget. Think Gerald R. Ford class aircraft carrier going to space and back, with 5000 ton cargo capacity to LEO, or in that neighborhood.
My math on a NTR Starship: (1 ton = 1000kg throughout)
Two-stage Starship mass budget: 5000 tons, assuming max payload.
Dry mass: 180 tons
Current payload to LEO: 100 tons (my math says closer to 155t but whatever)
Even if you assume that an NTR version would be more than 5x the dry mass (quite a bit more than your estimate)
NTR mass budget (unchanged) 5000 tons
NTR dry mass: 1000 tons
Resulting payload to LEO: 800 tons
Plus, the NTR equivalent would be SSTO, straight up and back, only the tankage needing a refill, possibly for years. Isp really matters a ton and it outweighs a lot of the admittedly significant concerns you raise. Same thrust, more than 5x the dry mass, and it's still enough to 8x the cargo.
And of course this is without getting too sci-fi, no salt water Zubrin rockets or Project Orion or Project Pluto nonsense. SSTO using air as a working fluid until it gets up over ~30km is another nice concept.
This is all so theoretical though, and many decades in the future in the best case scenario. Someone with your opinions arguing about this in the 1970s would've argued we'd have had all this long before now, yet we don't. It's not clear that any progress will actually be made in the coming decades either; no one is seriously working on it like SpaceX and Blue Origin are working on their Mars rockets.
Chemical rockets have had thousands of successful missions. NTRs not only have zero, they don't even have a single complete rocket that's ever been built. If you want to get to Mars any time soon, don't pin your hopes on NTRs. The people who are actually doing the thing certainly aren't.
This isn't taking into effect the added mass of the engine itself, though. You'd need heavy radiation shielding, heavy shielding around the reactor itself so that it wouldn't be blown to bits in the event of a launch explosion, etc. Plus, hydrogen is the least dense propellant so your fuel tank needs to be much bigger than e.g. the RP-1/LOX fuel tanks used by the Falcon 9, Saturn V first stage, etc. You end up in a much worst place than even hydrolox, because at least in hydrolox the liquid oxygen fuel tank at least can be relatively small. With NERVA it's only hydrogen.
So, once you take into account these factors (rather than looking at the raw efficiency of just the engine alone), it ends up not being as much of an obvious win.
> You'd need heavy radiation shielding, heavy shielding around the reactor itself so that it wouldn't be blown to bits in the event of a launch explosion, etc.
Not nearly as much as you think. You need some shielding between the reactor and any radiation-sensitive payloads (including humans) but you don’t need anything around the reactor; radiating into outer space is kind of like pouring water into the ocean.
This requires the reactor not to be operating while on the ground, of course, and probably not even fueled until it reaches orbit. (Note that I said “fuel”—ie uranium—and not “propellant”.) But since the purpose of NTR’s is to enable deep space travel beyond cislunar space, that’s not a blocker.
> Plus, hydrogen is the least dense propellant so your fuel tank needs to be much bigger than e.g. the RP-1/LOX fuel tanks
NTR’s are competing with hydrolox upper stage engines for propulsion in space. They aren’t competing with high-thrust RP-1 or methane engines for launch. In fact RP-1 isn’t even in the picture after you reach LEO. SpaceX is going with methane over hydrolox for Starship largely because liftoff from Mars is a requirement and methane can be synthesized on Mars.
> You end up in a much worst place than even hydrolox, because at least in hydrolox the liquid oxygen fuel tank at least can be relatively small. With NERVA it's only hydrogen.
This makes zero sense. There’s no reason you couldn’t use oxygen as propellant in an NTR; it’s just that hydrogen works better.
Again I think you’re missing the point—density is important for launching from the ground because during launch, you need to be able to produce a TWR > 1. But once you’re in orbit, none of that matters anymore.
Upper stages—especially ones for going anywhere past LEO—are already predominantly hydrolox because the better specific impulse of hydrolox, combined with its low mass, more than compensates for the added dry mass of tankage.
Oxygen in an NTR provides an ISP barely better than chemical rockets, and how reactive would 1500K oxygen be with your reactor? with the additional mass requirements of massively heavy NTR engines, shielding and cooling oxygen doesnt make sense.
We'll see ... personally I think the proof is in the pudding that despite having had these designs on the drawing board for 60+ years, there hasn't even been a single complete rocket that's ever materialized for any of them, let alone a successful mission. Contrast that with the thousands of successful missions using chemical rockets over that time period.
So it's not overstating it to say that there are some problems with nuclear thermal rockets, otherwise they would be commonplace by now.
It doesn't take an NTR to land on Mars. If you're going to pin all your hopes on NTRs I think it's still going to take decades. No one is even seriously working on them.
I'd much rather pin my hopes on chemical rockets in the form of SpaceX's Starship.
I didn't say it did. But it's one of the two most feasible options, the second involving lunar ISRU and orbital propellant depots. NASA's Mars Design Reference Mission has consistently included NTR studies.
Going past Mars and into the Belt, an NTR is practically essential.
> If you're going to pin all your hopes on NTRs I think it's still going to take decades.
Anything we try is going to take decades if you're talking about real time and not Elon time. But that's beside the point.
The point I was trying to make is that there has been approximately zero serious investment in human spaceflight beyond LEO once Apollo wrapped up. The Soviets decided they didn't want to land on the Moon after all, the US decided they would rather build a flying space truck that goes to LEO than build on Apollo, and everyone else spent decades just catching up. Since an NTR is only useful for interplanetary flight, no interplanetary flight means no need to develop NTR.
The reactor isn't going to be running during launch - so it's not going to be highly radioactive. Also I'd expect the reactor to be fairly small, dense and robust - I doubt if it would be "blown to bit" in the event of a problem with the rest of the launcher.
e.g. During the 1980 Damascus Titan missile accident the missile exploded underground in a bunker and the warhead was thrown a fair distance but was recovered relatively intact:
It is going to be highly radioactive at some point (when it's in use), so it does need shielding for that point.
It needs to be able to survive any kind of catastrophic detonation/break-up/crash landing you can think of, because the alternative is spreading a large amount of radioactive material directly into the atmosphere/onto the Earth's surface. For example, the Challenger orbiter itself was fine, but when the SRB blew up it took up the orbiter with it. So if there'd been any nuclear materials in the orbiter, even if they weren't used at launch, they still would've needed serious shielding.
And it doesn't matter if the consequences of said radioactive material release aren't actually that bad in the grand scheme of things (like Fukushima looks not to have been) -- what matters is that the public reaction to such an event would preclude the possibility of ever launching it again.
The nuclear fuel doesn't really need to be shielded any better during launch than the existing RTGs used on space missions. Scaremongering notwithstanding, those are pretty much impossible for a mere launch accident to "blow to bits" in a way that exposes the PuO, and so would the fuel rods of a proper reactor. At least unless the reactor itself suffers a catastrophic excursion and blows itself to bits à la Chernobyl No. 4, but we've become pretty good at building reactors that don't do that.
And the RTGs use an isotope that provides significant energy just from radioactive decay. By contrast, uranium in a reactor is barely radioactive at all, before the reactor has been turned on.
If they're launching it over the ocean (they would be, if they were launching from Kennedy or Vandenberg) they could just say "YOLO" and let the reactor fall into the ocean. It certainly wouldn't be the first time a nuclear reactor was dumped into the ocean (either by accident or deliberately.) It's obviously not a popular thing to do, but ocean water does provide a lot of shielding..
While true, a reactor blown up over land would also be simpler to clean up; while a reactor scattered over the ocean bed would be easier to leave in place.
"Simpler" here is relative. It would probably still cost billions of dollars and would create enough of a PR disaster to permanently forestall an NTR from ever launching again.
Water is not remotely as good as lead is by mass as a radiation shield. If you aren't already hauling many cubic meters of water for other reasons (and you wouldn't be), it absolutely wouldn't make sense to bring it along solely for that purpose. Water also has the severe problem of being liquid at habitable temperatures, so you either have to continuously spend energy to freeze it or deal with absolutely ruinous slosh.
The linked article is talking about the radiation shielding properties of what essentially amounts to a large swimming pool's worth of water. You know how much that would weigh?? Water is not a good radiation shield, it's just cheap here on Earth, so we use large quantities of it in applications where weight doesn't matter.
> It is going to be highly radioactive at some point (when it's in use), so it does need shielding for that point.
We can use two things as shielding - propellant and distance. For the first, we would want to make sure we don't use all the propellant with the reactor running hot, but decrease power output as we run out of propellant. As for distance, we may want to add a foldable structure between the propulsion section and the habitable section that would be extended to its full length prior to starting the nuclear reactor.
The long foldable structure would add considerable weight* and complexity (and thus risk), and would also be unsuitable for use as a first or second stage. So, this would maaaaybe be acceptable to handle the Hohmann transfers on a crewed mission to Mars, but not for much else.
* It needs to be strong enough to handle the full acceleration of the engines.
> It needs to be strong enough to handle the full acceleration of the engines.
I don't think we can expect to have high accelerations with NTRs anyway, so I don't think it'd need to be particularly robust. Also, propellant tank walls can be structural elements here too. If a foldable structure isn't practical, we could just assemble it from rigid elements lofted on other launches.
The problem is that Hall-effect thrusters provide basically no thrust.
Most Hall-effect thrusters provide millinewtons of thrust, with the best lab model producing 5.4N and the planned AEPS motor producing a theoretical max of 2.356N.
A single Merlin 1D motor produces 690kN of thrust at sea level. You would need over 290,000 AEPS engines to match one Merlin 1D engine. At 40kW per engine, this would use 11.7 gigawatts to run, so you’d also have to add in the weight for a dozen or so nuclear power plants.
Keep in mind that the falcon 9 uses nine Merlin 1D motors for the first stage.
> You would need over 290,000 AEPS engines to match one Merlin 1D engine.
Even worse, even an infinite number of AEPS engines would not match a single chemical rocket engine of any type, because Hall effect thrusters have a thrust-to-weight ratio of significantly less than 1. So it can't even lift itself off the pad, let alone anything else.
> So it can't even lift itself off the pad, let alone anything else.
This is only a problem for take off from large bodies. Mass per mass, you'll get a lot more delta-v from a Hall engine than from a Merlin. It'll just take a lot longer to get to that point.
Which brings up an interesting option for NTRs: bimodal propulsion. From a single propellant tank you could drive both an NTR and a Hall thruster, depending on your acceleration needs. Hall thrusters aren't heavy (certainly not compared to fission reactors) and you could use your NTRs when you needed high acceleration (such as to reduce transit times or to take off and land from a big rock) and use the low-power extreme-efficiency Halls for most of the trip when you'd be coasting.
> This is only a problem for take off from large bodies.
Yes, that's what we're talking about here.
> Mass per mass, you'll get a lot more delta-v from a Hall engine than from a Merlin.
Yes, that's what specific impulse measures (which is what most of the larger conversation has been about).
As for bimodal propulsion, it makes you wonder if that would be worth doing at all, or if it wouldn't be better to just have separate engines and fuel tanks so each can do what it's best at. I suspect it might end up being less weight and complexity just to use an ion thruster as designed using xenon propellant. Xenon fuel tanks aren't typically all that big anyway.
If I’m doing my math right, the 25kg AEPS engine would need to produce 245N to even hover without any payload. So it’s theoretical max is about 0.9% of what’s needed to even lift itself.
And of course these figures get much worse when you consider the other components that are a necessary part of the engine working, namely, the xenon fuel tank and massive solar panel array.
For a relevant comparison, the thrust-to-weight ratio of the Merlin 1D engine by itself is something like 180:1, but the overall thrust-to-weight ratio of the entire rocket at launch is only 1.4:1. So just in order to get a rocket that can lift off at all and deliver useful payload to orbit you need at least a 3-digit thrust-to-weight ratio on the engine itself. Ion thrusters don't remotely come close.
We're talking about orbit raising, not launch, which the nuclear thermal can't do. I can't believe I've been downvoted this much when I probably know more about electric propulsion and orbital dynamics than anyone in this thread.
Also, downvoted for not wanting to accidentally create radioactive fallout? Do the cost benefit analysis when there's plenty of other technologies for moving mass around off planet and to other planets.
Convenient calculator: http://www.quantumg.net/rocketeq.html