Gooey Pebble bed reactor

[KerikBalm]

So I was looking at atomic rockets the other day, and I had a reactor idea that I haven't heard of before. I guess it would have similarities to a nuclear lightbulb - a gas core in a solid enclosure.

I propose as a nearer tech reactor, a liquid core in a solid enclosure. Solid core designs are limited by the melting point of the fuel rods. Proposed liquid core and open cycle gas designs have problems with escaping fissile elements and their products, because the propellent is in contact with the gaseous or molten fissile material. The NERVA design had an exhaust temp of under 2700 K. I've seen some proposals to get that up to maybe around 3300 K... but... we have materials that can withstand much higher temperatures than that.

https://en.wikipedia.org/wiki/Tantalum_hafnium_carbide

Why not take a pebble bed reactor, enclose the pebbles in a Tantalum-Halnium carbide shell, and heat it up to.... oh, say about 4150 K? The alloy melts at about 4,263K, so this gives about 110C of a safety margin before the balls rupture and their internal uranium goo spills out and the engine spews out radioactive material.

Compared to the tested NERVA, this would be about sqrt(4150/2700) x higher Isp... 53% higher, which if the nerva was getting 800s, it would be about 1230s for the gooey pebble design, no?

But it should be even better, because above 3000K, you start to get H₂ disassociating into monatomic H, which gives a nice Isp boost. I'd imagine this would be significant by 4000K.

http://www.wag.caltech.edu/home/jsu/Thesis/node31.html

Thoughts? It seems to me like this would be a much shorter development cycle than liquid core or nuclear light-bulb designs, and would give pretty great numbers.

*edit*, just checked, Halfnium apparently absorbs neutrons very well. Tantalum has been looked at as a method of "salting" weapons, and when it captures a neutron it forms an isotope with a 114 day half life that emits gamma rays...

So these aren't the greatest materials to use in a nuclear core... hmmm

Edited September 25, 2019 by KerikBalm

[wumpus]

Is a magnetic (or otherwise virtual) throat and nozzle shown to be feasible?  I think the melting point of the materials that make these parts limit the Isp of NERVA, not the temperature of the reactor.

I have no clue about the workings of pebble beds to comment on those.

[Spacescifi]

1 hour ago, KerikBalm said:

So I was looking at atomic rockets the other day, and I had a reactor idea that I haven't heard of before. I guess it would have similarities to a nuclear lightbulb - a gas core in a solid enclosure.

I propose as a nearer tech reactor, a liquid core in a solid enclosure. Solid core designs are limited by the melting point of the fuel rods. Proposed liquid core and open cycle gas designs have problems with escaping fissile elements and their products, because the propellent is in contact with the gaseous or molten fissile material. The NERVA design had an exhaust temp of under 2700 K. I've seen some proposals to get that up to maybe around 3300 K... but... we have materials that can withstand much higher temperatures than that.

https://en.wikipedia.org/wiki/Tantalum_hafnium_carbide

Why not take a pebble bed reactor, enclose the pebbles in a Tantalum-Halnium carbide shell, and heat it up to.... oh, say about 4150 K? The alloy melts at about 4,263K, so this gives about 110C of a safety margin before the balls rupture and their internal uranium goo spills out and the engine spews out radioactive material.

Compared to the tested NERVA, this would be about sqrt(4150/2700) x higher Isp... 53% higher, which if the nerva was getting 800s, it would be about 1230s for the gooey pebble design, no?

But it should be even better, because above 3000K, you start to get H₂ disassociating into monatomic H, which gives a nice Isp boost. I'd imagine this would be significant by 4000K.

http://www.wag.caltech.edu/home/jsu/Thesis/node31.html

Thoughts? It seems to me like this would be a much shorter development cycle than liquid core or nuclear light-bulb designs, and would give pretty great numbers.

*edit*, just checked, Halfnium apparently absorbs neutrons very well. Tantalum has been looked at as a method of "salting" weapons, and when it captures a neutron it forms an isotope with a 114 day half life that emits gamma rays...

So these aren't the greatest materials to use in a nuclear core... hmmm

 

Physics does not make it easy, but I do applaud your research for trying to make a scifi drive that is achievable.

I think our materials science needs to catch up first.

Any idea what element we could engineer that has high heat loads but won't break up into radioactive compounds? 

Physics will likely have it's say there as well. Yet we will see what it says.

[sevenperforce]

6 hours ago, KerikBalm said:

So I was looking at atomic rockets the other day, and I had a reactor idea that I haven't heard of before. I guess it would have similarities to a nuclear lightbulb - a gas core in a solid enclosure.

I propose as a nearer tech reactor, a liquid core in a solid enclosure. Solid core designs are limited by the melting point of the fuel rods. Proposed liquid core and open cycle gas designs have problems with escaping fissile elements and their products, because the propellent is in contact with the gaseous or molten fissile material. The NERVA design had an exhaust temp of under 2700 K. I've seen some proposals to get that up to maybe around 3300 K... but... we have materials that can withstand much higher temperatures than that.

https://en.wikipedia.org/wiki/Tantalum_hafnium_carbide

Why not take a pebble bed reactor, enclose the pebbles in a Tantalum-Halnium carbide shell, and heat it up to.... oh, say about 4150 K? The alloy melts at about 4,263K, so this gives about 110C of a safety margin before the balls rupture and their internal uranium goo spills out and the engine spews out radioactive material.

Fantastic idea. As usual, great minds think alike. We discussed this very idea at length about two and a half years ago, both in this thread and on the CDE forums. Can't find the CDE forum thread, but we found that for SSTO applications, using a pebble-bed reactor with molten uranium inside of a tantalum halfnium carbide shell would be the ideal near-term design, most likely running on simple water for impulse density and ease of refueling. 

The only thing with higher performance that we could actually build without irradiating the world would be a nuclear lightbulb.

6 hours ago, wumpus said:

Is a magnetic (or otherwise virtual) throat and nozzle shown to be feasible?  I think the melting point of the materials that make these parts limit the Isp of NERVA, not the temperature of the reactor.

I have no clue about the workings of pebble beds to comment on those.

NERVA is absolutely performance-limited by reactor temperature. Get it too hot, and it melts. The nozzle and throat are actively cooled, just as in the SSME or F-1 or Raptor or Merlin, so they can handle any operating temperature. You can't actively-cool the reactor core because that's the thing that needs to be as hot as possible.

It's a common misconception that nuclear-thermal rocket engines are somehow unbelievably hotter than regular engines. They aren't. NERVA reached temperatures of around 3000K, while the SSME reached nearly 3600K. NERVA was more efficient not because it was hotter, but because the SSME required you to stuff a bunch of heavy oxygen atoms into the mix, while NERVA runs on pure hydrogen which is lightweight and thus achieves a higher specific impulse. This is also the reason that nuclear engines have shoddy T/W compared to any bipropellant chemical engine.

[kerbiloid]

Neutronic activation? Thermal expansion? Chemical compatibility with U?

[sevenperforce]

17 minutes ago, kerbiloid said:

Neutronic activation? Thermal expansion? Chemical compatibility with U?

Won't react with uranium, doesn't have significant thermal expansion. There's a little neutron activation and some neutron embrittlement but it's not significant; you'll need to reprocess your uranium before the TaHfC starts to cause problems.

[kerbiloid]

2 minutes ago, sevenperforce said:

doesn't have significant thermal expansion

Yes, actually they tested HfC-SiC ceramics for warhead nosecones and mini-fins(?). Just iirc they found it fragile, but probably this doesn't play role here.

Edited September 25, 2019 by kerbiloid

[KerikBalm]

14 hours ago, sevenperforce said:

Fantastic idea. As usual, great minds think alike. We discussed this very idea at length about two and a half years ago, both in this thread and on the CDE forums. Can't find the CDE forum thread, but we found that for SSTO applications, using a pebble-bed reactor with molten uranium inside of a tantalum halfnium carbide shell would be the ideal near-term design, most likely running on simple water for impulse density and ease of refueling.

Ah, I had missed the detail about the interior of the pebbles being molten. I thought it was a matter of the pebble's uranium being alloyed with something to increase melting point. I guess that detail didn't stick in my memory.

I guess the pebble shells can be pretty thin, and the nuclear properties of the elements won't be such a problem.

I guess its just a vapor core, but cooler... but I don't quite understand a vapor core:

http://www.projectrho.com/public_html/rocket/enginelist2.php#id--Nuclear_Thermal--Vapor_Core

Quote

It is basically a solid core NTR where the solid nuclear fuel elements are replaced by chambers filled with uranium²³⁵ tetrafluoride vapor. The engine is admirably compact with a nicely low critical mass, and an impressive thrust-to-weight ratio of 5-to-1. However the specific impulse / exhaust velocity is only slightly better than a solid core.

In other words, the system is not to be developed because it has fantastic performance, but because it will be an educational step to building a system that does.

The specific impulse is around 1,280 seconds and the internal temperature is between 6,000K and 8,000K.

The uranium fuel is kept physically separate from the hydrogen propellant, so the exhaust is not radioactive.

as you said in the other thread:

Quote

The difference between a vapor core rocket and a nuclear lightbulb is that the former uses heat transfer through the casing to get the propellant going, whereas the latter uses hard x-ray blackbody emissions from the gas core to heat up the propellant.

In the case of a vapor core or molten core, its still heat transfer through a solid casing, so you end up limited by how hot the outside of the casing can be... so what the heck is the advantage of a vapor core over a molten core? The inside of your casing needs to be below its melting point, so the temperature differential across the casing wall should be the same (even if the center of the gas core is hotter, it has to be cooler along the edges, or it melts the casing).

It seems to me vaporizing the core shouldn't given any advantage, and it just makes it more difficult to keep the case from melting. The limit should be the case melting point, no?

[sevenperforce]

2 hours ago, KerikBalm said:

Ah, I had missed the detail about the interior of the pebbles being molten. I thought it was a matter of the pebble's uranium being alloyed with something to increase melting point. I guess that detail didn't stick in my memory.

I guess the pebble shells can be pretty thin, and the nuclear properties of the elements won't be such a problem.

I guess its just a vapor core, but cooler... but I don't quite understand a vapor core:

http://www.projectrho.com/public_html/rocket/enginelist2.php#id--Nuclear_Thermal--Vapor_Core

In the case of a vapor core or molten core, its still heat transfer through a solid casing, so you end up limited by how hot the outside of the casing can be... so what the heck is the advantage of a vapor core over a molten core? The inside of your casing needs to be below its melting point, so the temperature differential across the casing wall should be the same (even if the center of the gas core is hotter, it has to be cooler along the edges, or it melts the casing).

It seems to me vaporizing the core shouldn't given any advantage, and it just makes it more difficult to keep the case from melting. The limit should be the case melting point, no?

You're right -- there aren't many advantages for a vapor core over a molten core, and the limit is the melting point of the case. There are, however, a few advantages. A vapor-core rocket uses uranium hexafluoride gas, which can be more easily removed for reprocessing than liquid. It's also lightweight in comparison to a molten-core NTR. As Project Rho explains, the benefit of building a vapor core is learning how to build a nuclear lightbulb, which is where the real magic happens.

[kerbiloid]

Spoiler

How many engineering physicists does it take to change a nuclear lightbulb?

 

[sevenperforce]

3 minutes ago, kerbiloid said:

  Hide contents

How many engineering physicists does it take to change a nuclear lightbulb?

 

Spoiler

Just three. One to change it into an open-cycle vapor core rocket, one to change that one into a fission-fragment rocket, and one to insist that we ought to go back to Project Orion.

A solid-casing molten core NTR has the same operating temperature as a vapor core rocket but it's heavier.

Most liquid-core designs, however, do not have solid casings. Rather, the molten fuel is centrifuged to keep it inside the thrust chamber while the liquid hydrogen is pumped through and heated by direct contact with the fuel. The outside of the chamber is actively cooled. This allows a much higher operating temperature than a gooey pebble bed or solid-casing molten-core rocket, but you do end up with uranium in your exhaust, which is generally not desired.

[kerbiloid]

3 minutes ago, sevenperforce said:

but you do end up with uranium in your exhaust, which is generally not desired.

Project Orion shyly smiles at the doors.

[wumpus]

19 hours ago, sevenperforce said:

It's a common misconception that nuclear-thermal rocket engines are somehow unbelievably hotter than regular engines. They aren't. NERVA reached temperatures of around 3000K, while the SSME reached nearly 3600K. NERVA was more efficient not because it was hotter, but because the SSME required you to stuff a bunch of heavy oxygen atoms into the mix, while NERVA runs on pure hydrogen which is lightweight and thus achieves a higher specific impulse. This is also the reason that nuclear engines have shoddy T/W compared to any bipropellant chemical engine.

I've heard a claim that the SSME was temperature limited*, although perhaps merely because they stopped the R&D when they could survive the needed temperature.  And of course the hottest part of the SSME will be the combustion gasses themselves: in NERVA the reactor has to be hotter than the hydrogen and has to warm that up (presumably from cryogenic temperatures, unless active cooling is needed elsewhere).  But now that you mention it, of course the nozzle will be cooler than the reactor (the throat might not be thanks to adiabatic heating, and might be tricky to cool actively).

*An astronaut giving a speech on youtube (possibly a TED program).  No idea how much it was dumbed down, but it was aimed at people who didn't know the rocket equation so take it with a grain of salt.  It still was pretty good, but you had to notice plenty of "lies to children" going on.

[sevenperforce]

9 minutes ago, wumpus said:

I've heard a claim that the SSME was temperature limited*, although perhaps merely because they stopped the R&D when they could survive the needed temperature.  And of course the hottest part of the SSME will be the combustion gasses themselves: in NERVA the reactor has to be hotter than the hydrogen and has to warm that up (presumably from cryogenic temperatures, unless active cooling is needed elsewhere).  But now that you mention it, of course the nozzle will be cooler than the reactor (the throat might not be thanks to adiabatic heating, and might be tricky to cool actively).

When hydrolox is burned at stoichiometric ratio, it burns notably (though not vastly) hotter than the maximum operating temperature of the SSME. The lower temperature used in the SSME did help keep the engine cooler and reduce stresses, but the primary reason for the lower temperature was to add more hydrogen to the propellant flow. By adding more LOX, the engine could have burned hotter (with additional, though not prohibitive, active cooling), but pushing less hydrogen at higher temperature would have resulted in lower bulk specific impulse than pushing more hydrogen at lower temperature.

[sevenperforce]

6 hours ago, kerbiloid said:

Project Orion shyly smiles at the doors.

You know, it just occurred to me – why do we need to have the “shell” be solid at all?

Build an actively-cooled cylindrical thrust chamber with a layer of uranium enclosing a layer of a less-dense, non-radioactive metal like hafnium or molybdenum, hollow in the center. Withdraw the control rods or add the reflectors (or however you want to do it) and let it start to heat up, while centrifuging the whole affair. Eventually, the uranium will melt. Then the hafnium will melt. However, because uranium is so much denser than hafnium, they will remain separated by the centrifugal forces and thus the uranium will never be in direct contact with the propellant flow through the center.

The reactor can run all the way up to the boiling point of uranium metal (7100 C) with all the heat going by direct thermal transfer straight into the hydrogen propellant. Critically, the propellant never touches the uranium and so there is no significant radioactivity in the exhaust, just like with a solid-core NERVA or a graphite pebble-bed.

V_e² ∝ R_gas*T_c and we know that a bare-metal liquid core at 3000 K has a specific impulse of around 1600, so this design would be able to push up to a specific impulse of 2500 seconds. Assuming you can get off the ground (a LOX-afterburner, water injection, and/or ejector shroud come to mind), you can reach LEO with margin to spare and a 32% fuel fraction.

[kerbiloid]

3 minutes ago, sevenperforce said:

boiling point of uranium metal (7100 C)

Got it.

Spoiler

 

[farmerben]

5 hours ago, sevenperforce said:

You know, it just occurred to me – why do we need to have the “shell” be solid at all?

Build an actively-cooled cylindrical thrust chamber with a layer of uranium enclosing a layer of a less-dense, non-radioactive metal like hafnium or molybdenum, hollow in the center. Withdraw the control rods or add the reflectors (or however you want to do it) and let it start to heat up, while centrifuging the whole affair. Eventually, the uranium will melt. Then the hafnium will melt. However, because uranium is so much denser than hafnium, they will remain separated by the centrifugal forces and thus the uranium will never be in direct contact with the propellant flow through the center.

 

That is creative thinking, I like it.

The radioactive pollution concern is mostly fission products rather than uranium itself.  Containing the fission products is more difficult.  

 

[kerbiloid]

At that rotation speed, won't the fission products stay inside as well, due to the centrifugal force?
Except helium, of course. But it's not radioactive.

Edited September 27, 2019 by kerbiloid

[wumpus]

On 9/26/2019 at 2:28 PM, sevenperforce said:

The reactor can run all the way up to the boiling point of uranium metal (7100 C) with all the heat going by direct thermal transfer straight into the hydrogen propellant. Critically, the propellant never touches the uranium and so there is no significant radioactivity in the exhaust, just like with a solid-core NERVA or a graphite pebble-bed.

Direct thermal transfer without touching?  Is there an intermediary that is cooled by the hydrogen and heated by the reactor but (mostly) non-radioactive?  I think once I left the magnetosphere (probably just the atmosphere) I'd be more than willing to have radioactive exhaust (and jettison any intermediary).

[sevenperforce]

14 hours ago, wumpus said:

Direct thermal transfer without touching?  Is there an intermediary that is cooled by the hydrogen and heated by the reactor but (mostly) non-radioactive?  I think once I left the magnetosphere (probably just the atmosphere) I'd be more than willing to have radioactive exhaust (and jettison any intermediary).

If you have molten uranium and molten hafnium/molybdenum in the same cylinder being centrifuged, the higher density of the uranium will force it to the edge while the barrier metal will "float" in an inner cylindrical layer. Same temperature. The uranium will heat the non-radioactive barrier metal; the barrier metal will heat the hydrogen.

I probably should have done a mockup.

[kerbiloid]

Is there a model/picture/diagram of fission products vertical distribution in a melted volume of uranium?

As in fact all contained elements, including both uranium and hafnium, will be distributed along the radial direction, with some exponent-like law, won't the fission product be distributed through both hafnium and uranium layers as well? Especially the isotopes lighter than hafnium.

Edited September 28, 2019 by kerbiloid

[sevenperforce]

12 minutes ago, kerbiloid said:

Is there a model/picture/diagram of fission products vertical distribution in a melted volume of uranium?

As in fact all contained elements, including both uranium and hafnium, will be distributed along the radial direction, with some exponent-like law, won't the fission product be distributed through both hafnium and uranium layers as well? Especially the isotopes lighter than hafnium.

I would have to look closely at the fission products of various uranium or plutonium reactor designs in order to get an idea of how lightweight the barrier metal would need to be.

It is a shame that there is no antonym of "dense" other than "lightweight" which is really an antonym of "heavy".

[kerbiloid]

1 hour ago, sevenperforce said:

It is a shame that there is no antonym of "dense" other than "lightweight" which is really an antonym of "heavy".

Indeed, I tried to find it in google.

Though, at such temperature all of them will be monoatomic ionized gas, so probably "lightweight" will be a synonym to "antonym to dense" as well.

Edited September 28, 2019 by kerbiloid

[KerikBalm]

I'm guessing your concept has far too little surface area relative to the mass of the thing, and thus its TWR would be abysmal. Pebble beds are good because they have a large surface area to pump propellent through. Solid core NTRs have many channels to drive up surface area.

Your molten seal would be a single cylinder per centrifuge, and the surface area for heating propellant would be pretty bad I think, poor TWR

[sevenperforce]

50 minutes ago, KerikBalm said:

I'm guessing your concept has far too little surface area relative to the mass of the thing, and thus its TWR would be abysmal. Pebble beds are good because they have a large surface area to pump propellent through. Solid core NTRs have many channels to drive up surface area.

Your molten seal would be a single cylinder per centrifuge, and the surface area for heating propellant would be pretty bad I think, poor TWR

Make it longer and bubble the reaction mass through the inert metal rather than just injecting at the center.