Curious about nuclear thermal propulsion

[Cheap Astronomy]

As I understand it, an effective nuclear thermal propulsion engine for a large spacecraft has to run very hot, potentially so hot it would melt/degrade if not for the constant flow of hydrogen propellant through it. So my question is can a nuclear thermal propulsion engine stop and start during spaceflight through a vacuum - or do you have to run it continuously to avoid it destroying itself? If it can stop and start how do you keep it cool when stopped? If you are doing it by cycling coolant how do you keep the coolant cool in space - radiators? Thanks, Steve

[Noclevername]

The reaction can be damped down by a neutron moderator when not thrusting, ideally.

[cjameshuff]

As I understand it, an effective nuclear thermal propulsion engine for a large spacecraft has to run very hot, potentially so hot it would melt/degrade if not for the constant flow of hydrogen propellant through it. So my question is can a nuclear thermal propulsion engine stop and start during spaceflight through a vacuum - or do you have to run it continuously to avoid it destroying itself? If it can stop and start how do you keep it cool when stopped? If you are doing it by cycling coolant how do you keep the coolant cool in space - radiators? Thanks, Steve

Nuclear thermal rockets are designed around nuclear reactors that use a controlled fission reaction, not passive radioactive decay. They are controlled the same way other reactors are, using components that reflect, absorb, and/or slow neutrons to sustain and regulate the chain reaction. One of their problems is they can't start up or shut down as quickly as other engines, but they produce almost no heat until they're started and decay heat of the fission products reduces to almost nothing within hours after shutting down.

[Van Rijn]

Cheap Astronomy, you might be thinking of the heat dissipation issue with commercial power reactors, but these are far smaller, more like test reactors or small power reactors, which, due to their much smaller volume, don’t have as much trouble rejecting heat. They can be inherently safe, not needing active cooling to deal with post-shutdown residual heat.

Also, Nerva and pebble bed designs have used graphite for much of the interior reactor structure which can deal with quite high temperatures.

[Noclevername]

Cheap Astronomy, I can recommend the Atomic Rockets website. It's set up for laymen to learn about, well, atomic rockets including nuclear-thermal engines and all kinds of others.

[Cheap Astronomy]

Thanks everyone - great information.

[cjameshuff]

Cheap Astronomy, you might be thinking of the heat dissipation issue with commercial power reactors, but these are far smaller, more like test reactors or small power reactors, which, due to their much smaller volume, don’t have as much trouble rejecting heat. They can be inherently safe, not needing active cooling to deal with post-shutdown residual heat.

Also, Nerva and pebble bed designs have used graphite for much of the interior reactor structure which can deal with quite high temperatures.

NTRs do need active cooling for residual heat. They produce ~6% of their normal output immediately after shutdown, and decay over an amount of time related to how long the reactor was in operation. This is much shorter than it is for groundside power reactors, but can still easily be hours, and cooling can require enough propellant that it's more mass-efficient to carry multiple NTR engines and jettison each after a single use.

[Noclevername]

So radiators would still be required, just not as large as with other nuclear options?

[cjameshuff]

So radiators would still be required, just not as large as with other nuclear options?

A bimodal NTR would need radiators, but realistically I think you're just going to use solar for power instead of trying to cram your core full of low-power heat exchangers without compromising its ability to heat propellant. Real NTR mission design will require bringing cooldown propellant or treating NTRs as expendable one-shot engines.

[Noclevername]

Solar can provide both thermal and electric options. Mix and match as desired, all needing low mass, large surface area reflectors or collectors. Inflatable frames to support lightweight graphene sheets, that sort of thing.

[kzb]

NTRs do need active cooling for residual heat. They produce ~6% of their normal output immediately after shutdown, and decay over an amount of time related to how long the reactor was in operation. This is much shorter than it is for groundside power reactors, but can still easily be hours, and cooling can require enough propellant that it's more mass-efficient to carry multiple NTR engines and jettison each after a single use.

Hang on though. The 1960's NTR's were designed as upper stages escaping Earth gravity, needing plenty of thrust As such they generated Gigawatts, and 6% of a GW is 60MW. a lot of heat.

The whole point of a reactor is to produce a high a temperature as feasible. So the thing will be engineered to withstand very high temperatures. The rate of heat loss is proportional to T^4 as is well known, so there is a very steep dependence of power radiated on the temperature.

For example at 1600K you need about 5 sq m radiating surface per MW of heat. Assuming it is two-sided that is a square radiator 1.6m on a side (for 1MW).

Reactors for interplanetary use, i.e from Earth orbit onwards, do not need to be very powerful. I dunno, 100MW probably plenty, and 6% of that is 6MW.

You then need c. 30 sq metres total radiating area. If your radiator is two sided it needs to be half of this, i.e 15 sq metres.

It's not a show stopper is it?

[cjameshuff]

Hang on though. The 1960's NTR's were designed as upper stages escaping Earth gravity, needing plenty of thrust As such they generated Gigawatts, and 6% of a GW is 60MW. a lot of heat.

The whole point of a reactor is to produce a high a temperature as feasible. So the thing will be engineered to withstand very high temperatures. The rate of heat loss is proportional to T^4 as is well known, so there is a very steep dependence of power radiated on the temperature.

For example at 1600K you need about 5 sq m radiating surface per MW of heat. Assuming it is two-sided that is a square radiator 1.6m on a side (for 1MW).

Reactors for interplanetary use, i.e from Earth orbit onwards, do not need to be very powerful. I dunno, 100MW probably plenty, and 6% of that is 6MW.

You then need c. 30 sq metres total radiating area. If your radiator is two sided it needs to be half of this, i.e 15 sq metres.

It's not a show stopper is it?

1600 K is a hair under the melting point of inconel, and well above the boiling point of sodium. You're talking about a system that could use liquid copper as a coolant. Protecting the rest of your spacecraft from its radiators would be a significant problem, never mind the core.

And a radiator's no good if you can't get heat to it. That requires heat exchangers in the core, in the way of getting propellant to "as high a temperature as feasible". A core you can cool with radiators will invariably perform worse as a NTR. You can use the same heat exchangers to produce power with the core running at lower temperature, but you would likely be able to get it from solar power for less mass unless you're far in the outer system.

I didn't say it was a showstopper, I said they need active cooling to avoid melting down. The most realistic method for that is to continue flowing propellant at a lower rate as the core cools.

[kzb]

1600 K is a hair under the melting point of inconel, and well above the boiling point of sodium. You're talking about a system that could use liquid copper as a coolant. Protecting the rest of your spacecraft from its radiators would be a significant problem, never mind the core.

And a radiator's no good if you can't get heat to it. That requires heat exchangers in the core, in the way of getting propellant to "as high a temperature as feasible". A core you can cool with radiators will invariably perform worse as a NTR. You can use the same heat exchangers to produce power with the core running at lower temperature, but you would likely be able to get it from solar power for less mass unless you're far in the outer system.

I didn't say it was a showstopper, I said they need active cooling to avoid melting down. The most realistic method for that is to continue flowing propellant at a lower rate as the core cools.

Yes 1600K is hot but what is the reactor temperature when it is operation? According Wikipedia this was 2406K, back in the 1960's.

It seems possible to me that, at this temperature, it would radiate a huge amount of heat without any special radiators or cooling system. Photos of the NERVA engines don't show any large radiators, although it has to be said they could've used continuing gas flow to cool on the stationary test rigs.

[cjameshuff]

Yes 1600K is hot but what is the reactor temperature when it is operation? According Wikipedia this was 2406K, back in the 1960's.

It seems possible to me that, at this temperature, it would radiate a huge amount of heat without any special radiators or cooling system. Photos of the NERVA engines don't show any large radiators, although it has to be said they could've used continuing gas flow to cool on the stationary test rigs.

...like I've been saying they'd need to do? You just quoted me going over the problems with radiators.

Not that a bimodal NTR would use its space radiators on a test stand anyway...these weren't flight systems, or even components of flight systems. NERVA apparently used liquid nitrogen for most of its cooling. An example full-power NERVA run from Wikipedia (apparently for 28 minutes, though it doesn't say so explicitly):

NRX A6 was started up again on 15 December. It ran at full power (1,125 MW) with a chamber temperature of over 2,270 K (2,000 °C) and pressure of 4,089 kilopascals (593.1 psi), and a flow rate of 32.7 kilograms per second (4,330 lb/min). It took 75.3 hours to cool the reactor with liquid nitrogen.

[kzb]

...like I've been saying they'd need to do? You just quoted me going over the problems with radiators.

Not that a bimodal NTR would use its space radiators on a test stand anyway...these weren't flight systems, or even components of flight systems. NERVA apparently used liquid nitrogen for most of its cooling. An example full-power NERVA run from Wikipedia (apparently for 28 minutes, though it doesn't say so explicitly):

I'm just saying it doesn't seem reasonable to dispose of such an expensive piece of kit when you could have, at worst, a few square metres of radiators instead.

BTW I found this pic for an updated nuclear rocket:


https://usnc.com/ultra-safe-nuclear-...esign-to-nasa/


I don't know if the fins are meant to be cooling fins, but if they are they're not very big.

[cjameshuff]

I'm just saying it doesn't seem reasonable to dispose of such an expensive piece of kit when you could have, at worst, a few square metres of radiators instead.

BTW I found this pic for an updated nuclear rocket:


https://usnc.com/ultra-safe-nuclear-...esign-to-nasa/


I don't know if the fins are meant to be cooling fins, but if they are they're not very big.

And again, it takes a lot more than a few square meters of radiators.

[publiusr]

Stan Borowski and Steve Howe are two individuals looking at nuclear thermal
VASMIR is nuclear electric, as was the JIMO concept. Fission fragment rockets and pulse Orion craft are nuclear concepts as well.
Nuclear Salt Water Rockets interest me the most—and I wonder if the concept of jacketed thrust might help prevent melting—perhaps in autophage rockets.