I’m still skeptical. Nuclear just has too many characteristics that limit its ability to make use of that ISP.
Starship will have a dry mass around 85 tons including about 9 tons of engines, payload of 100 tons, and 1,200 tons of methane/LOX fuel. At an ISP of 380 that gives it a deltaV of around 9.8 Km/sec.
A similar NTR based Starship with a 750 ISP requires a lot more dry mass. The engines will mass at least 100 tons just to provide 1/15th the thrust of Methane based Raptors. That’s ok in space, given you can just fire them longer but they will need hundreds of tons of radiators and shielding. And more mass to insulate the hydrogen tanks.
A dry mass of 500 tons would give the NTR a deltaV of 10.7 Km/sec. That’s better, but it comes at a price. Your NTR Starship can’t land on Mars, it doesn’t have enough thrust. So you need dedicated landers, increasing dry mass further. And it’s hydrogen leaks, so you lose some of that DeltaV on a long voyage. And you can’t use anything other than hydrogen, or your ISP drops precipitously, and you can’t use any reactive fuels that will corrode your engine.
NTR designs need dramatically higher thrust to Weight ratios to ever become competitive with chemical rockets, even in deep space.
> but they will need hundreds of tons of radiators and shielding.
NTRs do not need radiators, they can keep cool using the prop. And the mass of shielding is measured in hundreds of kg to single-digit tons, not hundreds of tons. (You can very effectively reduce your shielding needs by putting the engine on a spar.)
The rest of your concerns are still valid. NTRs seem like they would work better if you are going to asteroids, or other such targets that don't have an atmosphere you can use to brake with, but I do not see the appeal for trips to Mars, where a more compact atmospheric-capable ship can shed half the trip Δv using a heat shield.
You need heavy radiators or your entire ship quickly becomes uninhabitable when the engine is running. We’ve never built a NTR capable of shedding all its heat through its propellant. In fact, we’ve never operated one in space, only on Earth where cooling is trivial.
They stay hot after shutdown and you'll still need to cool them to avoid damage. If you use your propellant/coolant for that, the ISP will drop further.
You plan for that in your burn, essentially doing a very low-powered long tail in it. Yes, doing this will slightly reduce Isp, but assuming you can alter the size of the throat on your bell, this effect is just single-digit seconds on the entire burn, compared to the "hundreds of tons of radiators" the GP referred to.
A variable geometry nozzle would help a lot here, by keeping a lower volume of propellant to still be ejected at a high speed. May be a good idea after all.
Starship will have a dry mass around 85 tons including about 9 tons of engines, payload of 100 tons, and 1,200 tons of methane/LOX fuel. At an ISP of 380 that gives it a deltaV of around 9.8 Km/sec.
A similar NTR based Starship with a 750 ISP requires a lot more dry mass. The engines will mass at least 100 tons just to provide 1/15th the thrust of Methane based Raptors. That’s ok in space, given you can just fire them longer but they will need hundreds of tons of radiators and shielding. And more mass to insulate the hydrogen tanks.
A dry mass of 500 tons would give the NTR a deltaV of 10.7 Km/sec. That’s better, but it comes at a price. Your NTR Starship can’t land on Mars, it doesn’t have enough thrust. So you need dedicated landers, increasing dry mass further. And it’s hydrogen leaks, so you lose some of that DeltaV on a long voyage. And you can’t use anything other than hydrogen, or your ISP drops precipitously, and you can’t use any reactive fuels that will corrode your engine.
NTR designs need dramatically higher thrust to Weight ratios to ever become competitive with chemical rockets, even in deep space.