Will SpaceX use NASA Kilopower Nuclear Reactors?

[RotoSequence]

Specific impulse doesn't matter as much when each impulse releases more energy than an entire rocket.

For the kind of money such a programme would cost it has to offer a serious increase in Isp. NTR is estimated to cost 10s of $Bn for a 2x increase over chemical Isp.  Orion would be much more expensive give the safety precautions needed throughout the whole design, build and operating of the system.

That was an aside reference to Orion type nuclear pulse propulsion, with each pulse supplied by an atom bomb, rather than nuclear thermal rockets.

[Robotbeat]

Specific impulse still matters tremendously.

[john smith 19]

Specific impulse still matters tremendously.

WRT to this thread the relevant measure is additional sec of Isp (compared to Raptor)/$ invested by SpaceX.

On that basis a NTR tug would be a good deal (even at only double the Isp) if the whole package cost less than whole conventional fuel + consumables for a BFS flight with Raptors provided it either cut the transit time at the same payload (less consumables, more payload, depending on how closed loop the ECLSS is) or it allowed BFS to carry a lot more payload for the same journey time

The same would apply to high thrust ion drives/VASIMIR/Fission fragment/magic fairy dust systems.

An important slide wrt the thread title,

What's important for commercial missions is price of Kilopower units. If they cost $100m, no way in heck it will be affordable. Even $10 million is a lot for just 10kW of power.

Let's not get too far ahead of the game here. Kilopower is in testing right now. So
1) A full Mars design does not exist yet, although NASA and its contractors are pretty familiar with Martian surface conditions
2) NASA has already shifted it's baseline DRM 5.0 architecture to use multiple Kilopowers rather than a single 40Kw unit. This suggests they really want Kilopower to succeed. Kilopower is good for NASA. A nice convenient size for a lot of missions, with (claimed) design stretch  into the MW range.
3) IIRC on Earth nuclear capacity is roughly $1000/Kw, so by "Big Nuclear" standards that's a $10K build cost (fuel on Earth is an ongoing expense).
4) But this is Mars, so what's a fair multiplier is negotiable. 10x Earth build cost? 100x? 1000x? that negotiation has not even started.
5) Nuclear can operate anywhere, and under any conditions, without power reduction. 24/7/365/. These are attractive qualities for a power system.
6) The availability of Methane deposits in large quantities is (potentially) another major game changer for the balance of judging where to put a first settlement down. With it and a big enough PV array to bootstrap LOX production, large scale mfg becomes much more possible.

So people should not fall into any negative thinking yet.  The ability to add capacity in man portable (the 40Kw design needed it's only specialist transporter, which was the lasted single item of downmass for the DRM 5.0 architecture) units, and the consistent output, are valuable benefits.

An important slide wrt the thread title,

What's important for commercial missions is price of Kilopower units. If they cost $100m, no way in heck it will be affordable. Even $10 million is a lot for just 10kW of power.

Especially since many apps will require tens of mega-Watts... $10billion is laughable from commercial perspective.

We are a long way from NASA even giving SX a possible price for this.

[AncientU]

SpaceX or any other commercial company will not be able to afford a buy from NASA.  Anything.  Ever.

[TrevorMonty]

For Lunar polar ISRU operations, even single 1KW reactor plus batteries maybe all that is needed to keep equipment warm and alive few days a month without sunlight.

Production would be suspended during these dark periods.

[Jcc]

SpaceX or any other commercial company will not be able to afford a buy from NASA.  Anything.  Ever.

Then it's a good thing NASA doesn't really sell technology, but more likely license it or give it away freely.

The R&D that is going into Kilopower is something no commercial company could justify doing (maybe Lockheed Martin once they finish their "mini" fusion reactor .

[RotoSequence]

Specific impulse still matters tremendously.

~6000 newton seconds for a bomb powered pusher plate design; up to 100,000 for a Medusa type bomb-sail design.

[BeyondNERVA]

Hi! I'm new here! Kinda weird doing this as my first post, but I found this thread from my blog being linked above, so it feels a BIT less strange... I'm an astronuclear geek, and write the Beyond NERVA blog. I've dug into this system quite a bit, but not SpaceX so much.

I haven't heard anything about Elon looking at nuclear power in space, either fission power systems or nuclear propulsion, but it would make a lot of sense to. If you want NASA's take on the question of whether to do solar or fission, you can find it here: https://ntrs.nasa.gov/archive/nasa/casi.ntrs.nasa.gov/20160011275.pdf

(The other odd omission is a complete lack of any interest in Aldrin cyclers... which absolutely baffles me)

Space reactors are weird things... they kinda belong to everyone and no one at the same time, because they haven't had to be a thing. IIRC the fissile component is the responsibility of the DOE until integration into the full fission power system, when it gets handed over to NASA, but I'm not 100% sure on that. The design side tends to be a mish-mash of ideas and program requirements.

If you're interested in the origins of the program, I highly recommend Dave Poston's personal blog from before this program got started, during the Fission Surface Power program in 2010-2012: http://spacenuke.blogspot.com/

Someone posted my blog above about KRUSTY, if what you're looking for is not written up in there, check the in-line references: https://beyondnerva.wordpress.com/2017/11/19/krusty-first-of-a-new-breed-of-reactors-kilopower-part-ii/
if you're interested in the precursor fission test, DUFF, I did a writeup on that as well: https://beyondnerva.wordpress.com/2017/10/07/duff-father-of-krusty-kilopower-part-1/

There's talk of converting the core to low enriched uranium, with no insurmountable problems seen, there's just a lot that's different about this reactor, and the design particulars of nuclear spacecraft in general and this reactor in specific meant that the high security costs associated with HEU could be minimized. Basically, you take a 55%-74% mass hit on the full system if you do that, although there are areas that could possibly be optimized on the system. A recent paper by Dave Poston and Patrick McClure looks at it: https://fas.org/nuke/space/leu-reactor.pdf

If this design were to be commercialized (and it may be, BWXT could certainly handle it as a commercial provider - and they've got good connections at every point in the US nuclear supply chain), then it would probably be the LEU variant... but that will require a re-test of the core. Depending on how regulations are changed over the next few years (largely driven by advanced terrestrial designs, but space reactors will benefit as well), that could be either a very inexpensive test or virtually impossible to squeak through. Hopefully it's the former, and this team has done wonders on a shoestring and pocket change budget.

There are much larger variants of this reactor, which I look at briefly at the end of my KRUSTY rundown, called MegaPower. This is a Defense Nuclear Security Agency program, so you don't hear much about it, but it's rated up to 40 MWe, with a Brayton (?) PCS. My bet, though, is on a reworked version of the Fission Surface Power reactor, which is the next size class up from Kilopower, at 10 kWe - 1 MWe. It was the first fission system in this design series that proposed the Stirling PCS that I've seen developed to any degree, but it also had a very complex heat rejection system that ate the project's incredibly skimpy budget.

The design is basically solid, though, and could be reworked to overcome the problems that were seen during the development: https://ntrs.nasa.gov/archive/nasa/casi.ntrs.nasa.gov/20110007114.pdf

[jpo234]

I haven't heard anything about Elon looking at nuclear power in space, either fission power systems or nuclear propulsion, but it would make a lot of sense to. If you want NASA's take on the question of whether to do solar or fission, you can find it here: https://ntrs.nasa.gov/archive/nasa/casi.ntrs.nasa.gov/20160011275.pdf

We have a quote from Gwynne Shotwell during her talk at MIT from last September:

https://twitter.com/charlottelowey/status/913145922976190464?s=17

Shotwell on @SpaceX work on nuclear propulsion: "We're actually trying to get hold of some nuclear material - it's hard, by the way"

[jpo234]

I haven't heard anything about Elon looking at nuclear power in space, either fission power systems or nuclear propulsion, but it would make a lot of sense to. If you want NASA's take on the question of whether to do solar or fission, you can find it here: https://ntrs.nasa.gov/archive/nasa/casi.ntrs.nasa.gov/20160011275.pdf

We have a quote from Gwynne Shotwell during her talk at MIT from last September:

https://twitter.com/charlottelowey/status/913145922976190464?s=17

Shotwell on @SpaceX work on nuclear propulsion: "We're actually trying to get hold of some nuclear material - it's hard, by the way"

And we have a few quotes from the Tom Mueller Skype interview (https://zlsadesign.com/post/tom-mueller-interview-2017-05-02-transcription)

So we’re looking, actually, at like electric propulsion for the satellites, and we’re talking to people about nuclear-thermal, you know, the NASA centers are working on nuclear; it’s just prohibitively expensive to test because you can’t; it’s not like the 60s, like when you can just let fission products fly out of your rocket into the desert. You’ve now got to scrub it and clean it and capture it, which is super-expensive. I don’t think SpaceX could really afford to develop that rocket ourselves. If NASA ever gets turned on to develop those test stands, we’d probably want to jump in on that. You can just about double the performance of a rocket to Mars compared to a really-good, like a Raptor system, a chemical system, with fission; nuclear fission. Theoretically, fusion may be ten times better, and antimatter maybe a thousand times better, but I think those are certainly not going to happen in my lifetime. Maybe in your lifetimes.

It’s much better to use nuclear, fission reactor, it gets, you know, more compact; you actually get more; you get more power out per pound of reactor than you do out of solar cells, so it’s more mass-efficient. So if you’re taking it to Mars, it’s more efficient to ship reactors than it is to ship solar; it’s just that nobody’s really developed a space reactor yet. We’re working with NASA on that, and hopefully they’ll get funding to develop that. They’ve got a program called kilopower going that’s like, ten thousand watts, a 10 kilowatt reactor. We need a megawatt, but you know, you need to start somewhere.

[butters]

I'm confident that nuclear fission will be used for off-world surface power before it is used for orbital propulsion. Chemical propulsion is more "good-enough" than solar farms for Mars, and nuclear propulsion is barely superior at all for the Moon. We'll need reliable electrical power on Mars before the surface presence can grow to the point where nuclear propulsion really begins to pay off for supporting the supply chain.

[Robotbeat]

For Lunar polar ISRU operations, even single 1KW reactor plus batteries maybe all that is needed to keep equipment warm and alive few days a month without sunlight.

Production would be suspended during these dark periods.

That sort of thing is the best use of Kilopower, IMHO. Leaning on its strengths.

[TrevorMonty]

Low cost lunar and asteriod source fuel could eliminate need for nuclear propulsion. Especially for earth Mars trips.

The development cost of nuclear would pay for lot ISRU operations.
While ISRU fuel can compete against SEPs,  ISRU needs large scale solar power systems that are part of SEP development. 

[john smith 19]

Low cost lunar and asteriod source fuel could eliminate need for nuclear propulsion. Especially for earth Mars trips.

Provided you have a means to collect enough energy in the first place to process that raw material into propellant.

The development cost of nuclear would pay for lot ISRU operations.

Assuming they all were part of one unified budget.

But they are not.

And the point of Kilopower is that in fact it's budget has been very modest relative to previous plans to do this, which may explain why it's got this far, which is much further than earlier efforts ever have.

While ISRU fuel can compete against SEPs,  ISRU needs large scale solar power systems that are part of SEP development.

It's a balance.

[john smith 19]

There's talk of converting the core to low enriched uranium, with no insurmountable problems seen, there's just a lot that's different about this reactor, and the design particulars of nuclear spacecraft in general and this reactor in specific meant that the high security costs associated with HEU could be minimized. Basically, you take a 55%-74% mass hit on the full system if you do that, although there are areas that could possibly be optimized on the system. A recent paper by Dave Poston and Patrick McClure looks at it: https://fas.org/nuke/space/leu-reactor.pdf

That looks like a version of the Kilopwer architecture with LEU

If this design were to be commercialized (and it may be, BWXT could certainly handle it as a commercial provider - and they've got good connections at every point in the US nuclear supply chain), then it would probably be the LEU variant... but that will require a re-test of the core. Depending on how regulations are changed over the next few years (largely driven by advanced terrestrial designs, but space reactors will benefit as well), that could be either a very inexpensive test or virtually impossible to squeak through. Hopefully it's the former, and this team has done wonders on a shoestring and pocket change budget.

AFAIK BWXT is nothing to do with Kilopower, however it is much closer to the idea of "beyond NERVA," being an LEU NTR project, rather than an NEP (where I'm using the "P" for power, rather than propulsion).

There are much larger variants of this reactor, which I look at briefly at the end of my KRUSTY rundown, called MegaPower. This is a Defense Nuclear Security Agency program, so you don't hear much about it, but it's rated up to 40 MWe, with a Brayton (?) PCS. My bet, though, is on a reworked version of the Fission Surface Power reactor, which is the next size class up from Kilopower, at 10 kWe - 1 MWe. It was the first fission system in this design series that proposed the Stirling PCS that I've seen developed to any degree, but it also had a very complex heat rejection system that ate the project's incredibly skimpy budget.

The design is basically solid, though, and could be reworked to overcome the problems that were seen during the development: https://ntrs.nasa.gov/archive/nasa/casi.ntrs.nasa.gov/20110007114.pdf

Nuclear, even more so than space launch, seems obsessed with pedigree, the (traceable) history of a development.
So if Kilopwer can scale up with roughly the same materials and structure that's going to be viewed as the "less risky" option.  IIRC the increasing power output from the larger versions is mostly due to insertion of heat pipes inside the block, as opposed to just on the periphery.

WRT the Kilopower ground tests and the initial presentation I noted a 200c temperature drop due to poor conduction between two parts of the design.

Historically space thermal tactics have used carefully machined flat surfaces or equally carefully machined interlocking "hedgehogs" (like the cooling for the EMU's on the SSME's) to transfer heat.

However a small project  for the ECLSS on the ISS mentioned use of (essentially) carbon fibre knitted "socks" which when compressed between too surfaces could radically increase heat transfer (like thermally conductive grease, but with no danger of evaporation)

Aside from being more compact I thought HEU was easier for the DoE to procure, as it had quite a lot in stockpile from decommissioned nuclear weapons? It was (essentially) free.

[BeyondNERVA]

I haven't heard anything about Elon looking at nuclear power in space, either fission power systems or nuclear propulsion, but it would make a lot of sense to. If you want NASA's take on the question of whether to do solar or fission, you can find it here: https://ntrs.nasa.gov/archive/nasa/casi.ntrs.nasa.gov/20160011275.pdf

We have a quote from Gwynne Shotwell during her talk at MIT from last September:

https://twitter.com/charlottelowey/status/913145922976190464?s=17

Shotwell on @SpaceX work on nuclear propulsion: "We're actually trying to get hold of some nuclear material - it's hard, by the way"

And we have a few quotes from the Tom Mueller Skype interview (https://zlsadesign.com/post/tom-mueller-interview-2017-05-02-transcription)

So we’re looking, actually, at like electric propulsion for the satellites, and we’re talking to people about nuclear-thermal, you know, the NASA centers are working on nuclear; it’s just prohibitively expensive to test because you can’t; it’s not like the 60s, like when you can just let fission products fly out of your rocket into the desert. You’ve now got to scrub it and clean it and capture it, which is super-expensive. I don’t think SpaceX could really afford to develop that rocket ourselves. If NASA ever gets turned on to develop those test stands, we’d probably want to jump in on that. You can just about double the performance of a rocket to Mars compared to a really-good, like a Raptor system, a chemical system, with fission; nuclear fission. Theoretically, fusion may be ten times better, and antimatter maybe a thousand times better, but I think those are certainly not going to happen in my lifetime. Maybe in your lifetimes.

It’s much better to use nuclear, fission reactor, it gets, you know, more compact; you actually get more; you get more power out per pound of reactor than you do out of solar cells, so it’s more mass-efficient. So if you’re taking it to Mars, it’s more efficient to ship reactors than it is to ship solar; it’s just that nobody’s really developed a space reactor yet. We’re working with NASA on that, and hopefully they’ll get funding to develop that. They’ve got a program called kilopower going that’s like, ten thousand watts, a 10 kilowatt reactor. We need a megawatt, but you know, you need to start somewhere.

This is good news! (Minor nitpick, there have been 34 reactors flown, of three different designs, plus KRUSTY... and XE-PRIME was flight-qualified, so there have been reactors developed - and flown!)

The vast majority of the testing can be done with non-nuclear components, fortunately. A lot of the questions aren't that different from many chemical engines, especially LH2/LO2 engines. Most of the problems that Rover and NERVA faced were actually solved by the development of cryo hydrogen stages. The questions that remain tend to be on issues of thermal expansion, chemical reactions (2500+K hydrogen will ruin ANYONE's day if not handled properly), erosion and flow issues in the fuel elements... none of this study requires a nuclear reactor, just a rocket engine development facility with some specialized facilities.

These are still in the development phase right now, but I've got a page started for the types of test stands that have been built, and are currently operating. I hope to add more to it in the future, but... there's a ton of stuff out there that no one has had the time to make available outside conference papers, and test stands aren't sexy, neither are the engineering nitty-gritty details... for most people! Unfortunately, this is supposed to be a YouTube channel, the blog was just to point people to as kind of an in-depth FAQ that's taken on a life of its own fairly quickly.

https://beyondnerva.wordpress.com/nuclear-test-stands-and-equipment/

The reactor physics side has never really stopped, on a theoretical level. The problem is that no-one ever hears about it, because it's a small, specialized part of an attention-shy industry. When a lot of this work that I'm researching was done, maybe 1k people knew about it in any detail. That's changing now, but other than Winchell Chung at Atomic Rockets, there's no-one to repackage the information to make it more accessible, and there's WAY too much for just a couple people, or a hundred, to get the interested public caught up on 50+ years of materials science and technological development as it applies to in-space nuclear power.

Low cost lunar and asteriod source fuel could eliminate need for nuclear propulsion. Especially for earth Mars trips.

The development cost of nuclear would pay for lot ISRU operations.
While ISRU fuel can compete against SEPs,  ISRU needs large scale solar power systems that are part of SEP development.

I hear this a lot, and while it technically is true, it's not as efficient, or as powerful, as many designs that are possible with NTRs. Remember, we're playing in the kiddie end of the pool with what's possible here... a CERMET-fueled NTR will give you the same level of thrust for about 3x the isp of a hydralox stage, which is a very nice boost in capabilities. An open-cycle gas-core NTR, on the other hand, makes Hohmann transfers look like exactly what they are: pretty much the bare minimum effort to get anywhere.

The other advantage that an NTR has is it lends itself perfectly to having an afterburner. The LOX-augmented NTR uses a cascade injector ring to dump LOX into the hot H2 as it leaves the reactor and enters a combustion chamber. Your isp takes a tank, but your thrust gets a nice big boost.If you're trying to hit a particular launch window, that is a nifty little trick for a GNC to have up their sleeve.

This paper is now 20 years old, so the engine isn't exactly what we would try to build today, but the general concept is still just as valid.

https://ntrs.nasa.gov/archive/nasa/casi.ntrs.nasa.gov/19950005290.pdf

There's talk of converting the core to low enriched uranium, with no insurmountable problems seen, there's just a lot that's different about this reactor, and the design particulars of nuclear spacecraft in general and this reactor in specific meant that the high security costs associated with HEU could be minimized. Basically, you take a 55%-74% mass hit on the full system if you do that, although there are areas that could possibly be optimized on the system. A recent paper by Dave Poston and Patrick McClure looks at it: https://fas.org/nuke/space/leu-reactor.pdf

That looks like a version of the Kilopwer architecture with LEU

That's because it is. If you want to do the design commercially, this is the best way to do it in the short term - piggyback off difficult and expensive work someone else has already done!

AFAIK BWXT is nothing to do with Kilopower, however it is much closer to the idea of "beyond NERVA," being an LEU NTR project, rather than an NEP (where I'm using the "P" for power, rather than propulsion).

You're correct, but they also offer U-Mo fuel, of various types, and are able to do the same tooling and machining as Y12, who made this fuel element. Y12 is not a commercial enterprise, and the government can't sell anything, they need a commercial partner. BWXT is the logical choice. They already make all of the DOE's experimental fuel elements, most of the FEs for research reactors in the US, and fabricate, supply, and dispose of most (all?) of the US Navy's nuclear fuel as well. I can't think of anyone else even remotely as qualified...

Aside from being more compact I thought HEU was easier for the DoE to procure, as it had quite a lot in stockpile from decommissioned nuclear weapons? It was (essentially) free.

Absolutely. Y12 has unique procedures in regards to accountability of material (for good reason), which make HEU basically free in the overall operating budget (absolutely absurd...). This isn't an option for a commercial company, like SpaceX in the OP, who are stuck with LEU as long as they want to be an American company (or federal policy on that changes).

Nuclear, even more so than space launch, seems obsessed with pedigree, the (traceable) history of a development.
So if Kilopwer can scale up with roughly the same materials and structure that's going to be viewed as the "less risky" option.  IIRC the increasing power output from the larger versions is mostly due to insertion of heat pipes inside the block, as opposed to just on the periphery.

Not quite, the core also increases in size, which has a much bigger effect than you would think. a 200 MWt core and a 2000 MWt core of the same basic geometry are only marginally different in size. Nuclear scales UP very fast, but you tend to have a hard limit on DOWN fairly quickly... basically Flattop, which was the reactor used for DUFF.

WRT the Kilopower ground tests and the initial presentation I noted a 200c temperature drop due to poor conduction between two parts of the design.

Yeah, this was expected. Basically, it wasn't worth the money to re-tool, they'd just do conceptual work for the next iteration (which may or may not be a flight article). It's not a nuclear component, so GRC can do what they need to in order to fix the problem.

[john smith 19]

This paper is now 20 years old, so the engine isn't exactly what we would try to build today, but the general concept is still just as valid.

https://ntrs.nasa.gov/archive/nasa/casi.ntrs.nasa.gov/19950005290.pdf

Getting closer to 25 YO. And a belated welcome to the forum.

You're correct, but they also offer U-Mo fuel, of various types, and are able to do the same tooling and machining as Y12, who made this fuel element. Y12 is not a commercial enterprise, and the government can't sell anything, they need a commercial partner. BWXT is the logical choice. They already make all of the DOE's experimental fuel elements, most of the FEs for research reactors in the US, and fabricate, supply, and dispose of most (all?) of the US Navy's nuclear fuel as well. I can't think of anyone else even remotely as qualified...

Now that I did not know. They do sound like one of those outfits that's quietly been building their skills, and their relationship with the DoE.

Absolutely. Y12 has unique procedures in regards to accountability of material (for good reason), which make HEU basically free in the overall operating budget (absolutely absurd...). This isn't an option for a commercial company, like SpaceX in the OP, who are stuck with LEU as long as they want to be an American company (or federal policy on that changes).

I don't see either changing any time soon.

Not quite, the core also increases in size, which has a much bigger effect than you would think. a 200 MWt core and a 2000 MWt core of the same basic geometry are only marginally different in size. Nuclear scales UP very fast, but you tend to have a hard limit on DOWN fairly quickly... basically Flattop, which was the reactor used for DUFF.

For people used to conventional (LEU) reactor design these units are very small, but given you've not near minimum surface area and near maximum enrichment the only options left would be going fully enriched (100% U235), moving to a sphere (which looks a PITA to make and extract heat from) or a better reflector material(s). But I don't what is a better reflector, given this is a fast spectrum, rather than a thermal spectrum reactor.

WRT the Kilopower ground tests and the initial presentation I noted a 200c temperature drop due to poor conduction between two parts of the design.

Yeah, this was expected. Basically, it wasn't worth the money to re-tool, they'd just do conceptual work for the next iteration (which may or may not be a flight article). It's not a nuclear component, so GRC can do what they need to in order to fix the problem.

The Kilopower team have been very pragmatic regarding their of funds to test what really needs to be tested to demonstrate the viability of the concept, starting from DUFF, demonstrating the first use ever of a nuclear reactor to drive a Stirling engine.

BTW as I noted earlier large Stirlings are in commercial use for Diesel electric submarine propulsion. It's not done in the US, and it's not something you can get hold of easily, but it's certainly in the known SoA.

[BeyondNERVA]

And a belated welcome to the forum.

Thank you.

They do sound like one of those outfits that's quietly been building their skills, and their relationship with the DoE.

They helped build, and decommission, the USS Nautilus. IIRC they're one of the first commercial nuclear companies. They're also one of the more discrete, which has been appreciated during the anti-nuclear hullaboo of the last 40 years.

For people used to conventional (LEU) reactor design these units are very small, but given you've not near minimum surface area and near maximum enrichment the only options left would be going fully enriched (100% U235), moving to a sphere (which looks a PITA to make and extract heat from) or a better reflector material(s). But I don't what is a better reflector, given this is a fast spectrum, rather than a thermal spectrum reactor.

You're never going to want 100% enriched 235, it's an expensive and finicky pain in the butt that makes pretty much everyone nervous, and gives most people the heebie jeebies, for a reason. Having it be at 85+% is largely a holdover of working with fuel element geometries and critical assembly geometries that were originally designed for HEU, and are belatedly having LEU shoehorned into them.

I'm currently digging my way through documentation on the NCPS (Nuclear Cryogenic Propulsion Stage), which is one example of what I'm talking about. It started as a 95% enriched 235U, and is now currently being reduced to <20%, using CERMET fuels (https://beyondnerva.wordpress.com/2018/01/19/leu-ntp-part-two-cermet-fuel-nasas-path-to-nuclear-thermal-propulsion/). However, due to thermal constraints, propellant flow considerations, and the need to maintain a similar fuel element architecture in order to ensure the balance of the various elements and neutronic behaviors was correct in the reactor, the same ANL-2000 fuel element has been used throughout the program. Made out of different materials, with different enrichment, but the same fuel element nonetheless. This fundamentally limits the flexibility of the system, but at the same time this element has been tested in-reactor, and has data available that is unavailable on any other fuel element besides the graphite composite legacy NERVA fuel elements.

It should be relatively easy to work in a positive breeding ratio for the reactor, which would allow for the "useless" 238U can be bred into 239Pu, and then fissioned, without taking a significant mass hit... as long as you're willing to redesign your reactor from the ground up, including your fuel elements. Until 5-10 years ago, that idea was a non-starter. Combining discrete enough modeling for a full-flow expander cycle rocket engine, coupled with the same for a very high temperature gas cooled reactor, is still enough to give me the willies, but it's possible now, which is new. It doesn't replace testing, but hopefully KRUSTY will be that camel's nose in the tent that doesn't get the riding crop taken to it...

I expect we'll see lots of nifty things come down the pipeline in the next few years.

BeO is a good reflector in pretty much any spectrum. There are other options, and some quite interesting metamaterial options that have started peeking over the horizon, but those are still years away from an in-core test on the benchtop level.

BTW as I noted earlier large Stirlings are in commercial use for Diesel electric submarine propulsion. It's not done in the US, and it's not something you can get hold of easily, but it's certainly in the known SoA.

Very true. I guess I forgot to include the word "nuclear" in there...

[john smith 19]

They helped build, and decommission, the USS Nautilus. IIRC they're one of the first commercial nuclear companies. They're also one of the more discrete, which has been appreciated during the anti-nuclear hullaboo of the last 40 years. 

Would that have included any work on non-naval nuclear ships. Long term that would seem to be the way to go for Mars, unless bulk Methane mining work out.

You're never going to want 100% enriched 235, it's an expensive and finicky pain in the butt that makes pretty much everyone nervous, and gives most people the heebie jeebies, for a reason. Having it be at 85+% is largely a holdover of working with fuel element geometries and critical assembly geometries that were originally designed for HEU, and are belatedly having LEU shoehorned into them.

TBH none of them sounded that attractive, just possible options. Running with HEU was Kilopowers least attractive quality when looked at through the optics of public opinion about "The N word." 

I'm currently digging my way through documentation on the NCPS (Nuclear Cryogenic Propulsion Stage), which is one example of what I'm talking about. It started as a 95% enriched 235U, and is now currently being reduced to <20%, using CERMET fuels (https://beyondnerva.wordpress.com/2018/01/19/leu-ntp-part-two-cermet-fuel-nasas-path-to-nuclear-thermal-propulsion/). However, due to thermal constraints, propellant flow considerations, and the need to maintain a similar fuel element architecture in order to ensure the balance of the various elements and neutronic behaviors was correct in the reactor, the same ANL-2000 fuel element has been used throughout the program. Made out of different materials, with different enrichment, but the same fuel element nonetheless. This fundamentally limits the flexibility of the system, but at the same time this element has been tested in-reactor, and has data available that is unavailable on any other fuel element besides the graphite composite legacy NERVA fuel elements.

Don't underestimate that data, given the (historically) eyewatering cost of qualifying an element.
That's why I thought (if possible) a shared element between NTR and NEP would be a very good investment. Not optimal in performance, but cheaper than  2 separate qualifications and good enough to get the job done.

It should be relatively easy to work in a positive breeding ratio for the reactor, which would allow for the "useless" 238U can be bred into 239Pu, and then fissioned, without taking a significant mass hit... as long as you're willing to redesign your reactor from the ground up, including your fuel elements. Until 5-10 years ago, that idea was a non-starter. Combining discrete enough modeling for a full-flow expander cycle rocket engine, coupled with the same for a very high temperature gas cooled reactor, is still enough to give me the willies, but it's possible now, which is new. It doesn't replace testing, but hopefully KRUSTY will be that camel's nose in the tent that doesn't get the riding crop taken to it...

People make a big deal of how tough breeding is but most PWR have been breeding for decades in order to extend in core FE life.

I expect we'll see lots of nifty things come down the pipeline in the next few years.

At 10Kw you could power 2 QuitiQ T6 thrusters at 145mN at 4.5Kw each, and still have a Kw to run any science tasks en route to your destination.

BeO is a good reflector in pretty much any spectrum. There are other options, and some quite interesting metamaterial options that have started peeking over the horizon, but those are still years away from an in-core test on the benchtop level.

"Metamaterials?" That sounds very exotic for a reflector, or a moderator.  TBH for commercial projects I've always thought the best way to go would be natural Uranium. But that's tough.

Very true. I guess I forgot to include the word "nuclear" in there...

DUFF answered the question "Could you power a Stirling with a nuclear reactor" once and for all. that completely changed the debate from "We'd like to do this but no one's ever done it before" to "We don't have a Stirling linked reactor system available."

Hopefully KRUSTY will answer that second question in the next few months.

[BeyondNERVA]

Don't underestimate that data, given the (historically) eyewatering cost of qualifying an element.

Oh, I absolutely agree. The problem, though, is that in order to make modern NTRs, which meet the requirements of testability (and therefore the ability to reach flight qualification status) and LEU, we need to be able to test both new critical core geometries and in-core fuel element effects (mainly radiation and thermal, but erosive tendencies are the ultimate doom of a lot of FEs in tests). Back in the days of Rover, there were test stands to do this: Honeycomb for critical geometry experiments, and a modified Kiwi-A reactor for in-core cold flow testing were both at Los Alamos, and there were three reactors designed for repeated use with different fuel elements for hot fire testing (the A6-derived PAX, Pewee, and the Nuclear Furnace). That's the infrastructure that we need to rebuild, but I don't see that happening at Stennis. One design, maybe two, but not enough to develop a whole new fleet of FE designs.

That being said, I'm not a nuclear engineer, just a nerd with a bent for research, and a passion for the subject. I haven't heard much about a concerted push to a new fuel form, although there are some other designs floating around (that I'll also cover at some point this spring) that may be better options than NCPS. Those are also far, far less well researched, though, so once again the old FE geometry tends to win out on sheer cost.

That's why I thought (if possible) a shared element between NTR and NEP would be a very good investment. Not optimal in performance, but cheaper than  2 separate qualifications and good enough to get the job done.

I think that bimodal is the way to go. Depending on how power distribution in the core is done, it may be possible to do something like a KRUSTY PCS inserted into the propellant tubes, but that seems... iffy, I guess, especially at changeover. With bimodal, you've already got a secondary cooling system (primary is thermal propellant), PCS, and PCS cooling system in place. This is what the Russians are doing with the RD-0411, and it makes sense for so many reasons (I love that design, but I'm also an OTRAG fan - and they built a nuclear OTRAG!), not the least of which is being able to maintain hotel load for the spacecraft at minimum, while providing power for electric propulsion. You get the big dV kicks from the thermal side, then kick on the electric thrusters until orbital insertion.

The NTR is a specialized form of very high temperature gas cooled reactor. The GE 0710 fuel element (one of the two biggies in the CERMET world) was designed both for NTR and high temp gas cooled power plant use. Basically, you have a parallel to your system for delivering hydrogen to the propellant tubes, with a matching system to attach to the other side, for (usually) helium. If you want thermal, the hydrogen switches off, the hot-end He collector moves into place, and the hot He is run through a PCS (usually a Brayton). This is usually done at lower power, so radioactive flux is minimized (ALARA...), and often the limitation here is "how much mission mass do I want to waste on radiators?"

"Metamaterials?" That sounds very exotic for a reflector, or a moderator.  TBH for commercial projects I've always thought the best way to go would be natural Uranium. But that's tough.

77% mass hit, IIRC, for NU in Kilopower. It definitely can be done, though.

There are some things that are being played around with on the theoretical reactor materials side that kinda make my eyes pop. Biologically based gamma radiation shields are apparently an are of research in Russia (based on these: https://en.wikipedia.org/wiki/Radiotrophic_fungus), for instance, so the foil-and-cloth lined suits of 20 years from now may be enough to keep you completely safe, for quite high radiation fluxes, and ships (even inflatable ones) can be effectively shielded for reasonably low mass penalty. Unfortunately, I'm definitely NOT a materials guy (the CERMET post was about the limit for me), so I can nibble at the edges of some of these concepts but can't really explain it to myself, much less someone else!