2020's nuclear thermal propulsion efforts

[Asteroza]

I would say that the proposed development plan for the ESA NTER design is reasonably realistic, in the sense that the internals are essentially a classic NERVA NTR (thus not an advanced paper design, especially with the intended thermal margin allowing for reactor temps below even NERVA) with a turboinductor slapped on. If one is willing to accept the premise that classic NERVA NTR is a solved problem (but advanced variants, particularly with new fuel arrangements are not), then NTER is a solution that bridges most of the NTR performance gap that people commonly point out, in comparison to NEP systems. Unfortunately, being proposed by ESA means the necessary funding and facility availability is suspect, but the basic turboinductor "afterburner" concept is simple enough that anyone else can run with it. To be honest, I am surprised there isn't a russian variant of NTER at present (they seem to be pursuing a new NEP OTV tug).

[Proponent]

Assuming we're talking about Nuclear Thermal Rockets, there are still lots of gravity losses. Basically, the burn takes so long (sometimes between 30 minutes and nearly an hour) that you can no longer treat it as impulsive and so you lose much of the benefit of the Oberth Effect since your burn is not occurring as deep within the local gravity well.

Doesn't  that really depend on relative size of the engine thrust to the payload size?

Yes, but with the thrust-to-weight ratio of NTR being lousy to begin with, there is a high cost to pay for beefing up the acceleration.  Kirk Sorensen has written an informative blog post about this at Selenian Boondocks.

[RanulfC]

How about looking at NERVA engine designs. You won't see large radiator systems. I'm sure the engineers took heat into account.

I'm pretty sure NERVA derivatives baseline low flow propellant flushing for decay heat mitigation, AKA dribbling propellant through the reactor after the main burn. Functionally simple, but wasteful of propellant, and messes with your trajectory a little.

You'd be incorrect. NERVA had no post-thrust-flow to deal with decay heating as the engine was designed to handle the decay heat values. Point of fact during one test the propellant ran out 30 seconds early and the engine remained at full power output for a full minute after the propellant ran out. There was some slight melting on a few of the matrix rods but nothing that would have effected operations. The decay heat "issue" is only an issue to those who have not researched the NTR program.

The lack of any apparent radiator system for a system that was tested in atmosphere is not conclusive of what the system would have needed in space, this goes for EVERY device put into the vacuum of space, not just an engine.

Or it could be the fact the engineers who KNEW that the systems would operate in a vacuum and designed accordingly took such things into account. (Which they did) You assume that you are more cognizant of the "issues" with NTR design than people who have done the work for decades and a vastly deeper knowledge of the subject than you do, yet you feel qualified to question the work they've done?

Further more the expected use of NTR at the time would generally be single firing, TLI from LEO in which the engine along with the whole stage it was part of is just discarded as their was at the time or mentality of that day no such thing as reuse, refueling or any way to keep H2 liquid long enough to be relevant.  In this scenario decay-heat is irreverent and the whole stage can melt to slag for all you care.  After all many conventional engines were not restart-able so their was little demand for the issue to be solved, I have no doubt the issue and many others COULD be solved but it doesn't mean they been yet.

Again some, (at least a little possibly) research would save you some embarrassment here. The only NTR's assumed to be 'expendable' were specific ones suggested for a few Mars missions. The NERVA itself was designed from the beginning to be able to re-start multiple times per flight and was fully capable of 10 or more trips to the Moon and back to LEO before the reactor core began to decline in power. These "issues" were solved during the design phase and improved over the program test phase. NTR's had been tested for up to 2 hours with continuous operation up to 30 minutes, (due to the propellant storage limitations of the test facility) with multiple shut-downs and re-starts per mission. It was always meant to be part of a reusable system.

This pattern seems to come up frequently when people talk about NTR, they say it's all solved and we just need to build it with little or no development cost.  And when you point out that the performance isn't enough to justify it's use they point to advanced theoretical designs that were never tested.  Their is a consistent tendency to try to have the best of both and to compare what NTR's might be after BILLIONS in development to the present state of competitor techs while ignoring what those techs could become with that same spending.

If you want an NTR for current missions then everything IS solved we just need to qualify a newer, more modern engine. What you need to keep in mind is that when we talk of "off-the-shelf" engines we're mostly talking the stock NERVA using updated materials instead of things like graphite as a basis. We've designed and tested both ceramic and metallic cores but have not incorporated them into an actual engine. The flight-ready NERVA test article (XE-Prime) had a reactor power of 1137MW (thermal) and an ISP of 841s. NRX-A6 had a bit better with an average of 1199MW thermal and an ISP of over 860s but massed more.
Ceramic's could offer Isp's in the 900s and metallic cores might hit 1000s.

And that's before you add things like "bi-modal" where you use LOX injection to increase thrust for a short period use with ISP's still in the 800's, or "tri-modal" where you continue to run the reactor in a low-power mode and use an added power loop to provide electrical power which makes it easier to power up the reactor for burns.

The fundamental problem of course is the idea that the techs have to 'compete' for funding which they really don't. SEP has gotten far more money from commercial interests than from the government due to its more direct applications. NTR doesn't have such commercial applications (yet ) so has to depend on government programs to advance. S/NEP's are always going to be better for long haul missions with no time-constraints. Chemical will always have a place where high-thrust but short range is required and NTR could easily be slotted into the long-haul missions where time is a factor and higher thrust levels are required.

These currently funded developments look to me as nothing more then congress (or even specific Senators) trying to be an engineer, or a not so subtle way to subsidize research for the nuclear power industry.

Actually the nuclear industry has been researching more efficient power reactors for many decades and while there is some overlap it's only small since no commercial nuclear reactor would need to operate at the temps that an NTR would require. Yes the nuclear industry would find some benefit from improved NRT research, (the idea of using ceramic or metallic cores for example would elevate most of the dangers of a melt-down which is arguably a good thing to have) but it's far from a direct contribution.

As Ron notes you probably should do some research rather than just guess and assume

Randy

[john smith 19]

Actually the nuclear industry has been researching more efficient power reactors for many decades and while there is some overlap it's only small since no commercial nuclear reactor would need to operate at the temps that an NTR would require. Yes the nuclear industry would find some benefit from improved NRT research, (the idea of using ceramic or metallic cores for example would elevate most of the dangers of a melt-down which is arguably a good thing to have) but it's far from a direct contribution.
Randy

They have indeed but I'd say the the power generating industry is very reluctant to get involved.
They know enough about the actual economics of PWR's (rather than the promises that were made from the late 60's onwards) Vs conventional generation.
The licensing system is a bit like the launch range bureaucracy. They don't give allowances for smaller designs and I'm not sure they make allowances if failure modes are designed out. Likewise it looks (from my cursory reading) that a lot of this stuff is prescriptive to a PWR. So you've got rules specifically making sure a control rod can't be ejected by the 100 atm super heated water. Irrelevant if  you don't have a high pressure coolant to begin with. And so on.

However mostly OT to this thread.

I'm very dubious about NTR going anywhere, but Kilopower looks like it's making progress.

AFAIK NTR's (much like space power reactors) are so off the beaten path that there is limited cross over to conventional power generation.

[Asteroza]

AFAIK NTR's (much like space power reactors) are so off the beaten path that there is limited cross over to conventional power generation.

Most of the overlap is in high temperature coolant cycles, so the rejection temperature allows efficient air cooling for terrestrial reactors, as water cooling thermal powerplants has started to become an issue (exhaust water temperature limits for environmental reasons, such as water return temp when using river cooling). While the Heller cooling system was used at one nuclear plant in russia for an all air cooled system (that one was near the arctic circle), most reactors are in more temperate climates, which complicate things. There appears to be a going consensus that supercritical CO2 cooling cycles would allow pure air cooled nuclear plant operations (even for PWR's) much more easily, but the supercritical CO2 cycle technology stack (particularly the power turbine) has not yet fully matured in the the eyes of industry, and the stack has the most benefit when the reactor temperature is much higher than current PWR's (thus a lot of association with molten salt/LFTR systems as well as high temperature gas reactors). There's also some backend issues (will such high heat be exported as process heat for secondary customers, or stored for peaking purposes).

In the past space applications had helium coolant cycles, which had overlap with helium coolant/power cycles envisioned for pebble bed helium gas cooled reactors (particularly of german/south african origin, especially the TRISO pebbles), but a big issue was the design of helium power turbines, which is such a high jump from conventional air/nitrogen turbines. Supercritical CO2 is in that sense a middle ground between helium and air power turbines, but there is not a huge base of industrial knowledge backing either helium or supercritical CO2. There has been work at Sandia recently on supercritical coolant power cycles regarding the turbine (more molten salt reactor focused). and some practical work due to Allam cycle oxyfuel operations at an IGCC powerplant.

[john smith 19]

Most of the overlap is in high temperature coolant cycles, so the rejection temperature allows efficient air cooling for terrestrial reactors,

This is basic thermodynamics. I'm unaware of any design that uses hydrogen as a coolant. In terrestrial proposals nothing above 1000c either.

as water cooling thermal powerplants has started to become an issue (exhaust water temperature limits for environmental reasons, such as water return temp when using river cooling). While the Heller cooling system was used at one nuclear plant in russia for an all air cooled system (that one was near the arctic circle), most reactors are in more temperate climates, which complicate things. There appears to be a going consensus that supercritical CO2 cooling cycles would allow pure air cooled nuclear plant operations (even for PWR's) much more easily, but the supercritical CO2 cycle technology stack (particularly the power turbine) has not yet fully matured in the the eyes of industry,

I saw a report on a small CO2 turbine around 2 years ago. Nothing from then since. "Mature" implies you can place an order for one, pay money and get it delivered X months later. I'm aware of nothing like that.
If you can name a supplier I'd be very interested. Likewise the largest He turbine I'm aware of was 5MW, built in 1968 by the Germans for their high temp He pebble bed reactor design. Again nothing bigger built since.

and the stack has the most benefit when the reactor temperature is much higher than current PWR's (thus a lot of association with molten salt/LFTR systems as well as high temperature gas reactors). There's also some backend issues (will such high heat be exported as process heat for secondary customers, or stored for peaking purposes).

That's not an issue, that's a dream of advocates for such technology. To get that heat you'll need a suitable heat exchanger. The complete failure to talk about how they were going to do this left me highly suspicious of any claims about MSR's.

In the past space applications had helium coolant cycles, which had overlap with helium coolant/power cycles envisioned for pebble bed helium gas cooled reactors (particularly of german/south african origin, especially the TRISO pebbles),

Which ones? AFAIK actual space reactors have use molten metals and heat pipes. The proposals that have suggested Brayton cycles have IIRC been mixes of inert gases and (again IIRC) the biggest of those was 30Kw, not the 100s of MW (or multiple GW) So somewhere between 10 000 and 1 million times smaller than needed.

but a big issue was the design of helium power turbines, which is such a high jump from conventional air/nitrogen turbines. Supercritical CO2 is in that sense a middle ground between helium and air power turbines, but there is not a huge base of industrial knowledge backing either helium or supercritical CO2. There has been work at Sandia recently on supercritical coolant power cycles regarding the turbine (more molten salt reactor focused). and some practical work due to Allam cycle oxyfuel operations at an IGCC powerplant.

Practically none in fact. I saw the Allam reports dated 2018 but I've seen nothing more.

In fact AFAIK the only people looking to build real MW size He turbines Reaction Engines Limited. Everyone else is talking about what they could do.

But frankly all of this is quite OT WRT to the thread title. Space NTR systems nearest matching goals were the original PWR requirements IE Compact high power source for a submarine without much concern for thermal efficiency, but now including mass constraints. In fact an NTR has  zero energy recovery, discounting any picked up by the coolant to run the turbo pump. It's goal is to dump all its heat into its environment (accelerating the vehicle in the process). Hence you could say the whole design is the radiator.

[Lodrig]

Again some, (at least a little possibly) research would save you some embarrassment here. The only NTR's assumed to be 'expendable' were specific ones suggested for a few Mars missions. The NERVA itself was designed from the beginning to be able to re-start multiple times per flight and was fully capable of 10 or more trips to the Moon and back to LEO before the reactor core began to decline in power. These "issues" were solved during the design phase and improved over the program test phase. NTR's had been tested for up to 2 hours with continuous operation up to 30 minutes, (due to the propellant storage limitations of the test facility) with multiple shut-downs and re-starts per mission. It was always meant to be part of a reusable system.

I've done plenty of my own research and it consistently shows a gap between reality and hype.  We both know that reactor 'power' is not the issue in reusing the system, it's the physical integrity of the reactor, fuel cladding etc.  As for it being 'designed' to have a certain lifespan is not the same as having demonstrated and validated it, things are designed all the time which fail in testing, and testing in atmosphere is not a substitute for testing in vacuum.

If you want an NTR for current missions then everything IS solved we just need to qualify a newer, more modern engine. What you need to keep in mind is that when we talk of "off-the-shelf" engines we're mostly talking the stock NERVA using updated materials instead of things like graphite as a basis. We've designed and tested both ceramic and metallic cores but have not incorporated them into an actual engine. The flight-ready NERVA test article (XE-Prime) had a reactor power of 1137MW (thermal) and an ISP of 841s. NRX-A6 had a bit better with an average of 1199MW thermal and an ISP of over 860s but massed more.
Ceramic's could offer Isp's in the 900s and metallic cores might hit 1000s.

And that's before you add things like "bi-modal" where you use LOX injection to increase thrust for a short period use with ISP's still in the 800's, or "tri-modal" where you continue to run the reactor in a low-power mode and use an added power loop to provide electrical power which makes it easier to power up the reactor for burns.

The fundamental problem of course is the idea that the techs have to 'compete' for funding which they really don't. SEP has gotten far more money from commercial interests than from the government due to its more direct applications. NTR doesn't have such commercial applications (yet ) so has to depend on government programs to advance. S/NEP's are always going to be better for long haul missions with no time-constraints. Chemical will always have a place where high-thrust but short range is required and NTR could easily be slotted into the long-haul missions where time is a factor and higher thrust levels are required.

In earlier postings I point out how the 'off the shelf' performance  in the 800-900 range simply is not compelling enough to warrant using the technology over chemical rockets for the typical use scenarios like a mission to Mars.  The  best case designs back in the 90's for NTR powered Mars missions could barely get a 20-30% lower IMLEO over chemical due to the reactor and tank mass issues.  That's not enough to make an expensive tech development project attractive.

The case for NTR is getting worse all the time as costs of IMLEO fall and SEP systems grow ever more robust.  The middle ground your describing of 'long haul but high thrust' simply doesn't exist now if it ever did.  I challenge you to find a mission scenario in which NTR would beat out a Chemical, SEP or Chem/SEP hybrid system by enough to justify the development costs.

Lastly, all potential tech development absolutely dose compete for funding, both private and public.  I do not see how anyone can think differently.  NTR is just a continual loser when NASA engineers look at the potential benefits vs the costs.

[RanulfC]

I've done plenty of my own research and it consistently shows a gap between reality and hype.  We both know that reactor 'power' is not the issue in reusing the system, it's the physical integrity of the reactor, fuel cladding etc.  As for it being 'designed' to have a certain lifespan is not the same as having demonstrated and validated it, things are designed all the time which fail in testing, and testing in atmosphere is not a substitute for testing in vacuum.

I'd suggest that over 6 hours of operation at peak power show the physical elements are pretty sound. And yes 'designed' and (as mentioned) demonstrated again show differently. The research and testing was in fact heavily conservative AND taking into account the differences between atmosphere testing and vacuum operation. Your personal incredulity doesn't match professionals doing their job I'm afraid.

In earlier postings I point out how the 'off the shelf' performance  in the 800-900 range simply is not compelling enough to warrant using the technology over chemical rockets for the typical use scenarios like a mission to Mars.  The best case designs back in the 90's for NTR powered Mars missions could barely get a 20-30% lower IMLEO over chemical due to the reactor and tank mass issues. That's not enough to make an expensive tech development project attractive.

Oddly enough that seems to neglect that fact that many of those 90's NTR designs were used multiple times rather than 'thrown away' as the 70s missions were. Nor the fact they delivered more payload to Mars than a chemical design which often needed over 50% MORE IMLEO than the NTR mission for the same payload and transit time. As usually happens the reason that NTR gains interest is because it does in fact have advantages that prove attractive. Now of course the question is if that carries through in political and actual financial support.

The case for NTR is getting worse all the time as costs of IMLEO fall and SEP systems grow ever more robust.  The middle ground your describing of 'long haul but high thrust' simply doesn't exist now if it ever did.  I challenge you to find a mission scenario in which NTR would beat out a Chemical, SEP or Chem/SEP hybrid system by enough to justify the development costs.

Mars, Venus, Jupiter, Saturn pretty much anywhere and especially getting out of LEO to anywhere. Chemical, straight SEP or hybrid Chem/SEP almost always have longer travel times than NTR or NTR/SEP so it's less a question of "what missions" but of what the requirements are for the mission you plan. There's a good example here:
https://ntrs.nasa.gov/archive/nasa/casi.ntrs.nasa.gov/20120003776.pdf

And note it uses the smaller "PeeWee" class NTR's which DO in fact require shut-down cooling if they don't incorporate the electrical generation loop.

The development costs are really the question and it comes down to if you want a one-shot system or a long-term investment which requires more development. We can pretty much build a NERVA or PeeWee using modern materials fairly cheaply but it requires a test facility which no longer exists to actually run them. Of course we can always just loft them as you'd prefer and test them in space where they are meant to work anyway which would save a vast amount of money

Lastly, all potential tech development absolutely does compete for funding, both private and public.  I do not see how anyone can think differently. NTR is just a continual loser when NASA engineers look at the potential benefits vs the costs.

No all potential tech development does NOT in fact 'compete' for funding, as quite often happens funding comes from many sources to development that has enough interest which is why we're seeing NTR being considered again. EP systems have seen some pretty steady development for the last 40 years which is WHY it has improved so much. NTR on the other hand has barriers for private and other non-government funding so has been mostly theoretical and/or based on nuclear industry funding and work.

And even more odd is that you feel it is the NASA engineer's analysis of benefits and costs that has been the issue with NTR development as overwhelmingly NASA engineers FAVOR NTR development and have for decades. Pretty much exactly the opposite of your assertion. Where the problem is in political support and funding allowances, again mostly because the up-front costs and assumption of a long-term program have not appealed to politicians who in general don't care about long term space activity or work.

This is not new of course as politics has always been driven by short-term, focused goals with little effort to build sustainability or future utility. (One need look no further than the current "back-to-the-Moon-by-2024" by repeating Apollo with updated equipment driver to see this) The engineers are not the ones driving things because if they were they would have already significantly increased NTR research and development. (Along with a host of other development that has been stunted over the years but I digress )

Randy

[ThePonjaX]

Some news about DARPA request for a demostrator:

https://arstechnica.com/science/2020/06/the-us-military-is-getting-serious-about-nuclear-thermal-propulsion/

because seems Spacex is interested on nuclear propulsion:

https://twitter.com/charlottelowey/status/913145922976190464
Shotwell on @SpaceX work on nuclear propulsion: "We're actually trying to get hold of some nuclear material - it's hard, by the way"


https://twitter.com/elonmusk/status/1153378777893888005
Nuclear thermal rocket for fast transit around solar system would be a great area of research for @NASA

What do you think? We will see a proposal by SpaceX ?

[edzieba]

Can't see them doing more than partnering with another team to help with stage design. SpaceX have zero experience working with nuclear material in any capacity, and there haven't been any rumblings of them going on a hiring spree in the nuclear sector. I also can't see the DoD (or anyone else, for that matter) wanting to combine their "fail fast, fail often" development approach with "can we have some HEU to play with please?".

Then again, SpaceX have a habit of delivering surprises, and are getting a reputation for delivering period.

[RanulfC]

Elon Musk had NTR suggested to him at a conference and has been interested since he ran the numbers and like what he saw. He's aware though that the chance of SpaceX doing anything with it is low. Firstly as noted they have no experience and secondly it's more in line with government research and development than private sector work.

Their interest is both in propulsion and power btw since they see a need for both in the long term.

I have to admire how Ars Technica manages to screw up the whole history of nuclear rocketry in less than a single sentence though

Randy

[john smith 19]

Elon Musk had NTR suggested to him at a conference and has been interested since he ran the numbers and like what he saw. He's aware though that the chance of SpaceX doing anything with it is low. Firstly as noted they have no experience and secondly it's more in line with government research and development than private sector work.

Their interest is both in propulsion and power btw since they see a need for both in the long term.

I have to admire how Ars Technica manages to screw up the whole history of nuclear rocketry in less than a single sentence though

Randy

Well NASA seems to be fairly strongly behind kilopower so if they have a flight unit available by 2022 I'm sure SX could do some kind of deal. They provide the test environment (IE Mars) and get a copy of the test results.

DARPA OTOH is know for fairly low budge/high payoff research projects. Kilopower basically did this but I'm not sure well others will fit into this paradigm, given the power levels (GW Vs 100Kw-1MW at most? )

Should be interesting who puts in bids. If indeed anyone puts in bids. 

[RanulfC]

Well NASA seems to be fairly strongly behind kilopower so ifthey have a flight unit available by 2022 I'm sure SX could do some kind of deal. They provide the test environment (IE Mars) and get a copy of the test results.

DARPA OTOH is know for fairly low budge/high payoff research projects. Kilopower basically did this but I'm not sure well others will fit into this paradigm, given the power levels (GW Vs 100Kw-1MW at most? )

Should be interesting who puts in bids. If indeed anyone puts in bids. 

Well keep in mind this is DARPA and they are specifically looking at NTR for propulsion rather than surface power, (which is what NASA is looking at atm) so it's going to be a pretty niche bid. Then again this is what DAPRA does best even if their 'space' activity history is spotty. I'd be more surprised, (pleasantly mind you) if they manage to get permission to actually do a flight test.

Randy

[Lodrig]

Oddly enough that seems to neglect that fact that many of those 90's NTR designs were used multiple times rather than 'thrown away' as the 70s missions were. Nor the fact they delivered more payload to Mars than a chemical design which often needed over 50% MORE IMLEO than the NTR mission for the same payload and transit time. As usually happens the reason that NTR gains interest is because it does in fact have advantages that prove attractive. Now of course the question is if that carries through in political and actual financial support.

Your stating the exact same figures back to me just with a reversed ratio to make it sound bigger.  NTR being 33% less mass then Chem is the same as Chem being 50% greater then NTR.  My point still stands that it's not a compelling reduction in mass for the development and deployment of a new propulsion system.  To justify that kind of expense we generally need a system to be opening up whole new missions which were not even possible before, or offering order of magnitude reductions in launch mass for other missions.

Mars, Venus, Jupiter, Saturn pretty much anywhere and especially getting out of LEO to anywhere. Chemical, straight SEP or hybrid Chem/SEP almost always have longer travel times than NTR or NTR/SEP so it's less a question of "what missions" but of what the requirements are for the mission you plan. There's a good example here:
https://ntrs.nasa.gov/archive/nasa/casi.ntrs.nasa.gov/20120003776.pdf

No, just because a propulsion system is higher thrust doesn't mean it's automatically shorter travel time.  Inside Cis-lunar space and starting at LEO this will indeed be the case because any high trust injection that achieves Earth escape velocity or near too it will send a craft past the orbit of the moon in only a few days.  Over that length of time a low thrust system can't build enough speed to do much more then raise it's orbit.

But as soon as one departs earth orbit and begins traveling through Heliocentric orbits all this changes.  The transit times are inherently long now, long enough for low continuous thrust to equal or exceed the speeds of high thrust systems.

In addition any reasonable manned trip to another planet from Earth using a SEP vehicle would have that vehicle travel unmanned to a high earth orbit such as L2, with the crew taking a quick capsule ride to it only when it's ready to go.  This means that crew exposure to space environments can basically begin at the moment of Earth departure.  The same could be done for a NTR of course but it has no effect on travel time for the crew.

The paper you linked is indeed a good mission overview and point for comparison.  The outbound transit time of ~180 days is I believe a sweet-spot for human-health and a capability of a NTR, pushing transit times down to 150 days might be possible but would not be a good trade off for either NTR or current SEP systems.

In the DRM 5 supplemental, https://www.nasa.gov/sites/default/files/files/NASA-SP-2009-566-ADD2.pdf  (pages 28 to 29) they show that you can get a comparable travel time and IMLEO with a system power-density of around 20 kg/kw.  Though they assumed SEP would be at 30 kg/kw and NEP would be at 20 the two had identical ISP so effect of alpha is all that's represented here.  That's a reasonable near-term density and one that's a natural and incremental evolution of existing designs.

So again I can find no compelling case for NTR over SEP systems even for manned missions.

[john smith 19]

Well keep in mind this is DARPA and they are specifically looking at NTR for propulsion rather than surface power, (which is what NASA is looking at atm) so it's going to be a pretty niche bid. Then again this is what DAPRA does best even if their 'space' activity history is spotty. I'd be more surprised, (pleasantly mind you) if they manage to get permission to actually do a flight test.

Randy

Given that the actual missions this enables is classified that immediately limits the circle of potential bidders. 

DARPA was the final home of project Orion (as Freeman Dyson recalled in the article "Death of a Project"  )

Between the limited circle of bidders, the secret objectives list and the (likely) limited budget (not necessarily by DARPA standards, but by NTR standards. Wasn't the NTR line item in the NASA Mars DRM $133Bn?) this is going to call for some very creative engineering.   

The Killopower team could basically cast a 1 piece reactor core in HEU around a set of heat pipes and bolt on Stirling engines or simulators. Job done.

I'm not sure there's any design tricks like that for NTR. Electric pumps ("The full system will use LH2 cooled TE elements in the combustion chamber walls but as this is a PoC we just connect it to the mains")? Methane instead of LH2?  IMHO The big issue is testing.
a) Do it on the ground. But those test facilities don't exist anymore. How to recreate them affordably? Is a "baby" NTR (1-100MW?) even possible?
b) Forget that (too long to do, too expensive to build) and go straight for an on orbital test. Obviously one persons "Bold all-up test strategy" is another persons "Homicidal recklessness" so you'd need to make a rock solid safety case (no I don't have a clue how) but if you could.....

[RanulfC]

Well keep in mind this is DARPA and they are specifically looking at NTR for propulsion rather than surface power, (which is what NASA is looking at atm) so it's going to be a pretty niche bid. Then again this is what DAPRA does best even if their 'space' activity history is spotty. I'd be more surprised, (pleasantly mind you) if they manage to get permission to actually do a flight test.

Given that the actual missions this enables is classified that immediately limits the circle of potential bidders. 

THE problem with working on many DARPA projects is that the end-use and/or mission will be classified at the time of development work. You didn't have a huge pool of bidders anyway since NTR work has been limited to NASA centers and certain energy companies for decades. If DARPA can make the money inticing enough the bidders will come forward from the main industry R&D labs. What I'd LIKE to see is someone straight up propose the older DUMBO concept and get it on the llst
(An NTR engine with a known T/W of over 1/1 )

DARPA was the final home of project Orion (as Freeman Dyson recalled in the article "Death of a Project"  )

Yep, but then again a lot of the continued work on the pulse units was already there anyway... it would have been neat if they'd also gotten the infamous "Orion Battleship" model and been able to stick it in a warehouse somewhere

Between the limited circle of bidders, the secret objectives list and the (likely) limited budget (not necessarily by DARPA standards, but by NTR standards. Wasn't the NTR line item in the NASA Mars DRM $133Bn?) this is going to call for some very creative engineering.

Depends on what DARPA can get away with for requirements. If they require or are required themselves that a full up ground test take place prior to flight then it gets expensive as you'd need to build the test facility. On the other hand, in that case it might make SENSE for DARPA to get enough to put such a test facility together as a more 'neutral' source than say straight from NASA or the DoD. I doubt that will happen but we can hope.

The second option is if they are allowed to 'component' test but to defer a full up unit test until it's in deep space which is possible. In that case standard research and industrial "nuclear furnace" facilities can be used for most of the testing and full-up (but non-nuclear) ground test item can be used. Most of the 'work' will be simply updating and modernizing an existing and proven design, (likely the "PeeWee" or MITEE design) and integrating it into an existing upper stage. Mount it on a booster and launch it into an escape trajectory, fire it up and test it out. Again we can hope.

No it will likely be LH2 propellant as methane has a 'coking' (carbon build up) issue when used in an NTR and they are supposed to be 'prototyping' an "operational" design and that's likely the real propellant to be used. (And let's face it we have plenty of supporting and ancillary equipment that's off-the-shelf to use with it so it would make the most sense)

PeeWee, (here: https://en.wikipedia.org/wiki/Project_Rover, under "PeeWee") had a core that was only 53cm/21in, and MITEE (MinIature reacTor EnginE, who says scientist and grad students don't have a sense of humor ) is smaller and higher power.
(in fact the more I read the more it sounds like a VERY advanced DUMBO design)

See: https://www.osti.gov/servlets/purl/432864/

The MITEE core only masses 50kg/110lb but would have a thrust of 4,500lbf and in bi-modal mode a 100kw generation capability!

IMHO The big issue is testing.
a) Do it on the ground. But those test facilities don't exist anymore. How to recreate them affordably? Is a "baby" NTR (1-100MW?) even possible?

b) Forget that (too long to do, too expensive to build) and go straight for an on orbital test. Obviously one persons "Bold all-up test strategy" is another persons "Homicidal recklessness" so you'd need to make a rock solid safety case (no I don't have a clue how) but if you could.....

Yep pretty much but you can component test on the ground in current facilities and the size and power of the engine to be tested would drive the main costs of the full-up test facility. B is not so far fetched though as you can launch the test engine towards the Moon or deep space and only bring the reactor on-line once it's past the point of no return. If everything works you should have no problems maneuvering and changing trajectory, if it doesn't just let it impact the Moon or fly into solar orbit from the course it's already on.

Safety is in the design and the fact it's not radioactive until after the reactor is started and running.

Randy

[john smith 19]

THE problem with working on many DARPA projects is that the end-use and/or mission will be classified at the time of development work. You didn't have a huge pool of bidders anyway since NTR work has been limited to NASA centers and certain energy companies for decades. If DARPA can make the money inticing enough the bidders will come forward from the main industry R&D labs. What I'd LIKE to see is someone straight up propose the older DUMBO concept and get it on the llst
(An NTR engine with a known T/W of over 1/1 )

Di you mean 10/1? NERVA was about  > 7:1 from Sutton 4th Ed.

Yep, but then again a lot of the continued work on the pulse units was already there anyway... it would have been neat if they'd also gotten the infamous "Orion Battleship" model and been able to stick it in a warehouse somewhere

I suspect it's in someone's study and will turn up one day.

Depends on what DARPA can get away with for requirements. If they require or are required themselves that a full up ground test take place prior to flight then it gets expensive as you'd need to build the test facility. On the other hand, in that case it might make SENSE for DARPA to get enough to put such a test facility together as a more 'neutral' source than say straight from NASA or the DoD. I doubt that will happen but we can hope.

Well they did put up at least some of the cash for the Reaction Engines Inc high energy wind tunnel in Nevada, which I think is the first new US facility for some time.

The second option is if they are allowed to 'component' test but to defer a full up unit test until it's in deep space which is possible. In that case standard research and industrial "nuclear furnace" facilities can be used for most of the testing and full-up (but non-nuclear) ground test item can be used. Most of the 'work' will be simply updating and modernizing an existing and proven design, (likely the "PeeWee" or MITEE design) and integrating it into an existing upper stage. Mount it on a booster and launch it into an escape trajectory, fire it up and test it out. Again we can hope.

Now we're talking (I presume) a dedicated launch. Even just an F9 would swallow a big chunk of the budget. 

No it will likely be LH2 propellant as methane has a 'coking' (carbon build up) issue when used in an NTR and they are supposed to be 'prototyping' an "operational" design and that's likely the real propellant to be used. (And let's face it we have plenty of supporting and ancillary equipment that's off-the-shelf to use with it so it would make the most sense)

I don't have a real sense of where the costs are on such a programme or what facilities are still available. Baking RCC fuel elements seems kind of expensive.

PeeWee, (here: https://en.wikipedia.org/wiki/Project_Rover, under "PeeWee") had a core that was only 53cm/21in, and MITEE (MinIature reacTor EnginE, who says scientist and grad students don't have a sense of humor ) is smaller and higher power.
(in fact the more I read the more it sounds like a VERY advanced DUMBO design)

See: https://www.osti.gov/servlets/purl/432864/

The MITEE core only masses 50kg/110lb but would have a thrust of 4,500lbf and in bi-modal mode a 100kw generation capability!

Impressive.

Yep pretty much but you can component test on the ground in current facilities and the size and power of the engine to be tested would drive the main costs of the full-up test facility. B is not so far fetched though as you can launch the test engine towards the Moon or deep space and only bring the reactor on-line once it's past the point of no return. If everything works you should have no problems maneuvering and changing trajectory, if it doesn't just let it impact the Moon or fly into solar orbit from the course it's already on.

Safety is in the design and the fact it's not radioactive until after the reactor is started and running.

Randy

I had considered b) to be pretty far fetched.    It looks like to have any serious chance of success they winner bidder will have to be a) Highly motivated to succeed on a limited budget and b) really sharp in working out their DD&T strategy. Stuff they can build and test in existing ground facilities. Meshing that with a flight unit (not forgetting they'll need most of a stage. Maybe skip TVC by having a fixed nozzle and thrusters to point it in the right direction before main thrust is applied?)

Time will tell if this ends like Kilopower and the (potential) dawn of a new age of space nuclear power units, or the XS-1 programme. 

[RanulfC]

Did you mean 10/1? NERVA was about  > 7:1 from Sutton 4th Ed.

Erhm, Sutton being Sutton he seems to have lumped ALL nuclear engines under NERVA even though "Project Rover" has more than one NERVA design The basic "NERVA" engine topped out, (with a shield and all flight equipment added) at around 0.79-something to 1. NERVA derivative and some of the more advanced designs were expected to hit maybe 3.05 to 3.7 to 1 with LANTR for example. The PeeWee NERVA engine was tested and again assuming all the flight equipment and the final design systems that were never installed before the program went under and came out to 3.5 to 4.5 to 1. MITEE in it's most basic form was designed to hit 7.2 to 1. DUMBO on the other hand was designed to hit averages over 70 to 1.

I suspect it's in someone's study and will turn up one day.

Not likely! The few pictures show it's about the size of an office desk

Well they did put up at least some of the cash for the Reaction Engines Inc high energy wind tunnel in Nevada, which I think is the first new US facility for some time.

The US has been needing a new wing tunnel for a while so it was DARPA that got the attention but the general DoD that also contributed a lot.

Now we're talking (I presume) a dedicated launch. Even just an F9 would swallow a big chunk of the budget.

It really depends on how and who they get to support and/or partner with. The DoD could agree to flat out pay for it through Space Force as it's arguable to have a 'requirement' for the research. For that matter they may go whole hog and offer a Delta IV or other launcher and throw in buying a Centaur stage as a test-bed. A lot depends on who DARPA can rope into bidding and doing the actual research and construction.

I don't have a real sense of where the costs are on such a programme or what facilities are still available. Baking RCC fuel elements seems kind of expensive.

Metallic and CerMet (Ceramic/Metal) composite fuel elements are made and tested all the time and have been for decades. The issue is making and assembling enough into a working reactor. CerMet reactor elements based on NERVA and PeeWee were on the drawing boards when the program was shut down. Since then elements have been tested but not a reactor design and there is some differences between a carbon element design and the CerMet and/or metallic reactors but nothing that can't be done with what we know of reactor design.

We have several "nuclear furnace" facilities but they can only test small components scattered around but while there are some smaller stuff still available in Idaho what we really need is a reactor building and test facility with a scrubbed exhaust test stand. But DARPA can't afford any of that so would have to depend on current industrial and government facilities.

I had considered b) to be pretty far fetched.  It looks like to have any serious chance of success they winner bidder will have to be a) Highly motivated to succeed on a limited budget and b) really sharp in working out their DD&T strategy. Stuff they can build and test in existing ground facilities. Meshing that with a flight unit (not forgetting they'll need most of a stage. Maybe skip TVC by having a fixed nozzle and thrusters to point it in the right direction before main thrust is applied?)

Time will tell if this ends like Kilopower and the (potential) dawn of a new age of space nuclear power units, or the XS-1 programme. 

You can likely build and 'test' a reactor suitable with facilities at hand but actually running it as an NTR is problematical, hence the idea of "simply" replacing an RL-10 with the NTR on a Centaur

Oddly this comes at a time when the nuclear industry is lobbying for some significant changes to the US general policy on nuclear power. Part of this is being fed up with Congress over the lack of action on nuclear 'waste' in the form of "spent" power rods since part-and-parcel of the US nuclear industry is that they weren't given a choice on what to do with them since that was a responsibility taken on by the Federal Government. Literally the policy is that the government will gather, ship and store the used rods and that the industry just has to 'temporarily' store them till the government can collect them.

We've been storing them "temporarily" in the facility cooling ponds for almost 60 years now. Congress has never authorized the transportation of the rods, rejecting every proposed method and system. Yucca mountain has never opened up officially and frankly it would ONLY be used to store fuel rods at this point as every OTHER form of nuclear waste has had transportation systems approved and other storage facilities built and operated by private companies. (And note these transport and storage facilities are moving and storing stuff that's worse than the fuel rods on a daily basis)

The main problem is what is IN the fuel rods which is extract able amounts of plutonium. The choice in the 70s was between reprocessing the fuel rods, (which is what every other nation chose) and storing the rods as a way to deal with the plutonium. The US chose storage but never put into place the means or systems to carry out the policy they'd decided on.

The secondary problem is the current regulatory environment makes building new nuclear plants almost impossible not especially from a regulator stance but from the difficulty of getting any significant change in design approved due to the regulations being so focused on pressurized water reactors which limits implementation of more modern designs since they don't 'meet' the regulations wording but actually exceed the spirit of the wording.

Randy

[Llian Rhydderch]

Podcast on this subject, not listen to it yet.

http://fiso.spiritastro.net/telecon/Kokan-Joyner_6-3-20/

Interesting, looks like AJR is pivoting to NTR.

This is interesting.  Principally because the most consistent thing that the space industrial complex has been good at over six decades is getting funding out of the US Congress for programs that don't necessarily have high efficiency for the US taxpayer, some of which do little to advance space access.

If AJR and its partners were to be able to get substantial funding for NTR, the benefit to the advancement of the space frontier would be far in excess of the tens of billions spent so far on SLS, which is unlikely to ever have real efficiency for human space access.

[john smith 19]

Erhm, Sutton being Sutton he seems to have lumped ALL nuclear engines under NERVA even though "Project Rover" has more than one NERVA design The basic "NERVA" engine topped out, (with a shield and all flight equipment added) at around 0.79-something to 1. NERVA derivative and some of the more advanced designs were expected to hit maybe 3.05 to 3.7 to 1 with LANTR for example. The PeeWee NERVA engine was tested and again assuming all the flight equipment and the final design systems that were never installed before the program went under and came out to 3.5 to 4.5 to 1. MITEE in it's most basic form was designed to hit 7.2 to 1. DUMBO on the other hand was designed to hit averages over 70 to 1.

Ahh. I was looking at Sutton 4th Ed.  I hadn't realized quite how summarized the entry in the table was.
70:1 is much more like conventional rocket engines, and very impressive for an LH2 engine anyway.

Not likely! The few pictures show it's about the size of an office desk

There's an old story about a tire company that gets an enquiry for a rare kind of tire and they tell the enquirer "Sorry we can't supply it. It's being used by one of the managers as their office." 

The US has been needing a new wing tunnel for a while so it was DARPA that got the attention but the general DoD that also contributed a lot.

I'd never considered the possibility that while DARPA's name is on the check it could be funneling money from other parts of the DoD.

It really depends on how and who they get to support and/or partner with. The DoD could agree to flat out pay for it through Space Force as it's arguable to have a 'requirement' for the research. For that matter they may go whole hog and offer a Delta IV or other launcher and throw in buying a Centaur stage as a test-bed. A lot depends on who DARPA can rope into bidding and doing the actual research and construction.

Are they still making DIV's?  Definitely big $$. The guys behind Kilopower were pretty good at doing this as I recall. I can only hope the project management is equally motivated.

Metallic and CerMet (Ceramic/Metal) composite fuel elements are made and tested all the time and have been for decades. The issue is making and assembling enough into a working reactor. CerMet reactor elements based on NERVA and PeeWee were on the drawing boards when the program was shut down. Since then elements have been tested but not a reactor design and there is some differences between a carbon element design and the CerMet and/or metallic reactors but nothing that can't be done with what we know of reactor design.

It's a side point but since all weapons testing has been banned I'm guessing that nuclear explosion simulation has gotten a lot better, which suggests reactor design/simulation should also have much better definition.

We have several "nuclear furnace" facilities but they can only test small components scattered around but while there are some smaller stuff still available in Idaho what we really need is a reactor building and test facility with a scrubbed exhaust test stand. But DARPA can't afford any of that so would have to depend on current industrial and government facilities.

I always liked Reaction Engines approach. Use a single example of a full size module to avoid endless issues caused by scale effects (either known or unknown).  Are those facilities big enough to do that? That would just leave the final assembly (which of course has been extensively simulated already, and can now be anchored with the results from the actual sized parts).

But we are back to the test stand.  Also the size of such an installation will set the playing field for US NTR research for decades to come.   That suggests a minimum of an RL10 sized capacity.  But even that may just be too expensive.

Oddly this comes at a time when the nuclear industry is lobbying for some significant changes to the US general policy on nuclear power. Part of this is being fed up with Congress over the lack of action on nuclear 'waste' in the form of "spent" power rods since part-and-parcel of the US nuclear industry is that they weren't given a choice on what to do with them since that was a responsibility taken on by the Federal Government. Literally the policy is that the government will gather, ship and store the used rods and that the industry just has to 'temporarily' store them till the government can collect them.

We've been storing them "temporarily" in the facility cooling ponds for almost 60 years now. Congress has never authorized the transportation of the rods, rejecting every proposed method and system.

I did not know this.    So the operators literally can't move them anywhere?  Hmm.

Yucca mountain has never opened up officially and frankly it would ONLY be used to store fuel rods at this point as every OTHER form of nuclear waste has had transportation systems approved and other storage facilities built and operated by private companies. (And note these transport and storage facilities are moving and storing stuff that's worse than the fuel rods on a daily basis)

The main problem is what is IN the fuel rods which is extract able amounts of plutonium. The choice in the 70s was between reprocessing the fuel rods, (which is what every other nation chose) and storing the rods as a way to deal with the plutonium. The US chose storage but never put into place the means or systems to carry out the policy they'd decided on.

No.

The main problem is that the reprocessing routes available back then originated with nuclear bomb making. It is no accident that they are very good at pulling Pu out. "Proliferation" (IE being able to build a Pu bomb) was a goal.  People have tweaked it over the years, but it's like putting a Porsche engine on a sit down mower. 
I've been doing a bit of digging on nuclear power and waste and what stunned me is that of every tonne of U put in about 955Kg is unchanged. [EDIT Perhaps we should start calling it Stored Nuclear Fuel, since it's is very far from "spent" ].
Maybe 10Kg are actinides (all fissionable but need hot cell processing. Pu being the majority by about 8.5Kg). All the real  trouble (I, Tc, Cs, Zr) are about 2Kg with Sr and Cs (30Yr 1/2 lives but lots of heat) about another 2Kg.

The other trouble is all the fluids conventional processes generate. Then I found a patent that was looking at zone refining SNF in the 1950's inside the fuel rod itself    It was talking about metallic U but there are ways to process metallic oxides into metal (even casting it)  without consumables. Various reports have described actual ZR runs on both U and Pu from ORNL and others since the 60's. Key issues seem to be stopping the skin oxidizing. Do that and Cs and some other FP's boil off the surface rather than accumulate at the ends, so keep the cell >30c and have  a good filtration system and that's one less problem to worry about.

Once you realize how tiny the amount of Pu is Vs the raw Uranium it is obvious the smart move is snip the U from everything (including the Pu) and process that in an environment fairly similar to the factory (obviously on a much smaller scale) where the pellets were originally made. 

The other actinides are more of an issue (Curium especially) but a much smaller volume. Keeping them all together means anyone trying to "divert" them will either die in the attempt or need very heavy shielding to cope with neutron emissions.

At that point you either have radically reduced the volume of waste you're storing on site or you need to find a way to split the actinides out.

Ideally you don't want their emissions to breed more U235/Pu from U238 so you want them to be put into some kind of Inert Matrix Fuel. ZrH seems a good choice (I didn't know but ZrO, which is another candidate IMF has worse thermal conductivity than U02  which is already pretty bad) but likely to give free H2 when it's processed in the next cycle. One little wrinkle would be to stack the Actinide pellets alternating with regular pellets. That exposes them to the unmoderated neutron spectrum from the ends of the regular pellets (roughly the first 10micro metres, judging by fission fragment rocket studies). Obviously tracking exactly which rods are regular U and which are reprocessed hot ones calls for tight inventory control, but I think a modern database system should be up to the job.  How many items can Amazon track through its warehouses in real time?

A truly first class system would also  strip out the Cs137 and Sr90 (c 30y half lives) and the stable isotopes (lanthanides and others) that make up about 30Kg IE the vast bulk of the burnt fuel. That just leave the hard core nasties of I129, Tc99, Zr93 and Cs 135.

A chart with actinides, short lived Sr and Cs stripped out shows Yucca could hold 225x  its current projected volume.

To non Americans your SNF management process makes no sense.   
Reading up on the background President Carter seemed like a good man with good intentions but this is definitely one for the "Law of unintended consequences" sample manual.   

The secondary problem is the current regulatory environment makes building new nuclear plants almost impossible not especially from a regulator stance but from the difficulty of getting any significant change in design approved due to the regulations being so focused on pressurized water reactors which limits implementation of more modern designs since they don't 'meet' the regulations wording but actually exceed the spirit of the wording.

Randy

Oh yes. Stuff like ensuring the control rods won't be ejected by the coolant pressure. What happens if the coolant is at 1atm, because it's a liquid metal,  and actually separate from the channels the control rods move in? Or what about if someone proposes control by liquid Cadmium eutectic alloy (say 140c) held in a tank above the reactor by gas pressure and warm by reactor heat?
Release gas, liquid fills vertical pipe. Instant shut down. Extremely fast acting and fail safe, but never seen before. OMG A reactor with no control rods. We can't license that. 

BTW I must recommend the OECD reports in this area. They are quite clear and give good discussions on all aspects of nuclear engineering.