NTR from the Ground Up

GoogleNaut · Guru

NTR from the Ground Up

G R L Cowan · Veteran Member

Lose the hydrogen.

Put a low-enriched or unenriched core on the bottom of a supercritical helium tank. At 3 K and 35 bar, helium density is 0.18 kg/L versus, for 15-K liquid hydrogen at the same pressure, about 0.079.

Helium from a 3-kK core should exit a rocket nozzle at 6.36 km/s if H2 from the same setup would exit, undissociated, at 9 km/s. Getting to orbit, 9.5 km/s, would then require mass ratios 4.449 for helium, the vehicle being 0.775 propellant by mass, and 2.874 for hydrogen, initial propellant mass fraction 0.652. Much smaller tanks on the helium rocket.

How does that work out ... for the same dry mass, the helium tank must hold (3449/1874), 1.84 times, as much fluid mass as the hydrogen tank. The stuff has (79/180) times, 0.44 times, as much volume per unit mass, so, OK, the tanks end up 81 percent as voluminous. 19 percent smaller. That's much, for small values of much.

A core surrounded by in-swirling cold hydrogen puts fission neutrons into it a distance inversely proportional to the free proton's 3.926-barn scattering cross section for such neutrons, and it very promptly thermalizes them. But then, it turns out, the free proton's scattering cross section for thermal neutrons is 20.47 barns. So each drunkard's step they take to get back to the fuel is 5.2 times shorter than the jump they made in leaving it. Since hydrogen-1's thermal neutron capture cross section is near 0.3 barns, they don't make it back.

(If the hydrogen in which they are thermalized is much below room temperature, the situation is ... well, essentially the same. Even shorter steps, so a much tinier fraction can drunkard's-walk the distance a fission neutron jumps, but the complement of both fractions is close enough to 1.)

The situation is reversed with helium: fission neutrons jump short distances, thermal neutrons jump long ones. And if they had to do many thermal jumps, they still wouldn't be captured. So a nuclear thermal helium rocket can be very effectively moderated by its own propellant.

The fuel can therefore be, as above said, of low enrichment or even none at all, so if it falls into the sea -- coals to Newcastle.

With helium, there are some technical hitches with lox augmentation that make it a complication that can be done well enough without. So lose the oxygen too.

Let the baby light matches in the fuel room

GoogleNaut · Guru

Helium doesn't make any sense at all--you can't augment at all with oxygen because helium doesn't burn. Hydrogen burns with oxygen, which gives you the an energy source to drive the turbopump systems. Hydrogen is also two orders of magnitude cheaper to use than liquid helium.

Also hydrogen is totaly gassified by the time it hits the core--between regenerative cooling of the nozzle and the reflectors, it is a warm gas by the time it hits the core. There shouldn't be any phase transition of coolant in the core--so no real thermal shock issues there.

The thermal doppler effect of the core neutronics is interesting, but becomes less important once you operate the reactor at a fast spectrum--the speed of the neutrons is far, far higher than the speed of the reaction mass passing through the coolant channels, so tighter controls over reactivity are possible just with core geometry, placement of burnable poisons and control drums. Also, you have to have highly enriched uranium, otherwise the reactor will be a Pressurized Water Reactor that will weigh 2000 tons, instead of 10 tons. The highly enriched uranium is diluted within the tungsten matrix--this is the essence of concept of the solid solution--also this increases the melting temperature much higher than native uranium metal, but only slightly lower than tungsten. Also, a carefully designed reactor core will not have any pathological reactivity immersed in sea water like its smaller US NAvy cousins: insertertable poison rods can 'safe' the reactor into an essentially inert situation.

G R L Cowan · Veteran Member

Some of what 'Googlenaut' says must be in response to some posting other than mine.

Would it be a bad thing if a highly enriched core that had fallen onto the seabed were to continue cooking there, and be harder to approach, and detectable by a thermal plume?

A highly enriched core that had safed itself might turn out to have done so for the benefit of salvagers not otherwise associated with the project.

With an unenriched core, as I said, coals to Newcastle. It would intrinsically be 'safed': in ordinary water there is no way to 'unsafe' such a core. Sodium and chlorine would, I guess, make this also true for a slightly enriched core.

There is 4 billion tonnes of unenriched uranium in the sea. The only people who would much want a lost lump of it back would be the project principals, hoping to gain clues why it ended up there.

How shall the car gain nuclear cachet?

GoogleNaut · Guru

The trouble with an unenriched core is that the core must then be moderated--which adds weight and bulk to the core. For compactness sake, the core needs to be enriched. Every NERVA, Rover, Kiwi, and Phoebus reactor used enriched uranium...the US Navy has extensive experience with highly enriched uranium in reactors, and so on...

The 93% enriched uranium 235 is 'cut' with Tungsten to dilute it down so that its volume percentage is around 20%--low enough to not be usable as a weapon, but high enough to have a very high volumetric rate of energy production.

The whole idea with encapsulating individual reactor/engines is so that they can be quickly, easily recovered in the event of a problem. A launch abort will have a fairly well defined 'splash down' corridor along which recovery vessels could be prepositioned. A recovery capsule could be designed with floatation in mind, with parachutes, radio, optical, and acoustic recovery aids. Even something as simple as a 'satphone' making a call to the responsible orginizations reporting exact GPS splashdown cooridinates--all sorts of technological solutions are possible with today's recovery technology---many things not possible during Project Apollo.

Also, Tungsten is pretty stable in sea water. Not quite as stable as solid stainless steel or titanium, but for a couple of days in sea water it should be just fine. The US Navy has used Tungsten-Uranium alloys in its reactor cores for a long time.

Also, the whole idea with augmenting with liquid oxygen is the thrust 'amplification.' Isp isn't everything--thrust is also important. And high thrust at take off is most important of all--more important that very high Isp and Low thrust...

G R L Cowan · Veteran Member

GoogleNaut included:

... Every NERVA, Rover, Kiwi, and Phoebus reactor used enriched uranium...

... encapsulating individual reactor/engines ... so that they can be quickly, easily recovered in the event of a problem. A launch abort will have a fairly well defined 'splash down' corridor along which recovery vessels could be prepositioned. A recovery capsule could be designed with floatation in mind, with parachutes, radio, optical, and acoustic recovery aids. Even something as simple as a 'satphone' making a call to the responsible orginizations reporting exact GPS splashdown cooridinates--all sorts of technological solutions are possible ...

I seem to recall there was a really powerful safety rule that was also followed? What was that rule, again. Maybe you can dredge it up for the forum.

How shall driving gain nuclear cachet?

GoogleNaut · Guru

I'm not sure what you meant by a 'safety rule.' There were no NTR reactors ever flown--safety procedures would have been for the ground tests--and for those I'm sure they monitored the atmospheric plume because no scrubbers were used then...

As far as now, I think this does open up the discussion as to possible flight safety rules that might actually be stated for a launch...

I once proposed that reactor fuel (assembled cores) could be ferried up to an orbital assembly area using a 'boiler plate' CEV that was in effect just an aeroshell and splash down hull. The idea was to give a reactor core the same 'all aspect' abort survivability from a booster malfunction that astronauts have...also this gives the possibility of not just recovering the nuclear material, but the fully intact and undamaged payload for a later launch attempt. This is the kind of philosophy that I intended: treat the nuclear payload with the same 'respect' and accompanying rules as if the payload were a human being.

In that sense, then launching an NTR with nuclear engines running from the ground up, it behooves us to think of ways and means to recover intact the nuclear thermal engines in the event of a launch vehicle failure.

It would be best to think in terms of survivability to design the engines from the start with this in mind. For safety's sake, and with respect to responsibility to the environment, it makes sense to address the safety related issues concerning 1) engine erosion. Reducing erosion to as negligeable a level as can be reasonably and cost effectively managed. 2) Safety/survivability--identifying criticality 1 (failure of which can cause total vehicle loss: LOM/LOC/Dispersal of radionucleides) systems and improving their reliability to the point were catastrophic failures occur at the same relative rate as man rated systems. 3) Isolating catastrophic damage--if a Criticality 1 failure occurs, design for survivability. Turbopumps are 'blast isolated' from the reactor core. Engine has 'instant off' capability if a burn through is detected, turbopump fails (blows apart,) etc. Design criteria should address the possibility of an inflight scram and manage emergency heat dissapation with the goal of preventing dispersal of radio nucleides...That's a real tough one--I'm not sure if that can be even done---YET! And yet, these are precisely the kinds issues that need to be addressed before we can entertain the notion of an NTR from the Ground Up. From these basic philosphies, with the engineering realities of an actual design, can come actual flight rules. And right now, that's a pretty tall order.

-- Edited by GoogleNaut at 00:41, 2008-04-05

G R L Cowan · Veteran Member

GoogleNaut wrote:

I'm not sure what you meant by a 'safety rule.' There were no NTR reactors ever flown--safety procedures would have been for the ground tests--and for those I'm sure they monitored the atmospheric plume because no scrubbers were used then...

As far as now, I think this does open up the discussion as to possible flight safety rules that might actually be stated for a launch...

I once proposed that reactor fuel (assembled cores) could be ferried up to an orbital assembly area using a 'boiler plate' CEV that was in effect just an aeroshell and splash down hull. The idea was to give a reactor core the same 'all aspect' abort survivability from a booster malfunction that astronauts have...also this gives the possibility of not just recovering the nuclear material, but the fully intact and undamaged payload for a later launch attempt. This is the kind of philosophy that I intended: treat the nuclear payload with the same 'respect' and accompanying rules as if the payload were a human being.

In that sense, then launching an NTR with nuclear engines running from the ground up, it behooves us to think of ways and means to recover intact the nuclear thermal engines in the event of a launch vehicle failure.

It would be best to think in terms of survivability to design the engines from the start with this in mind. For safety's sake, and with respect to responsibility to the environment, it makes sense to address the safety related issues concerning 1) engine erosion. Reducing erosion to as negligeable a level as can be reasonably and cost effectively managed. 2) Safety/survivability--identifying criticality 1 (failure of which can cause total vehicle loss: LOM/LOC/Dispersal of radionucleides) systems and improving their reliability to the point were catastrophic failures occur at the same relative rate as man rated systems. 3) Isolating catastrophic damage--if a Criticality 1 failure occurs, design for survivability. Turbopumps are 'blast isolated' from the reactor core. Engine has 'instant off' capability if a burn through is detected, turbopump fails (blows apart,) etc. Design criteria should address the possibility of an inflight scram and manage emergency heat dissapation with the goal of preventing dispersal of radio nucleides...That's a real tough one--I'm not sure if that can be even done---YET! And yet, these are precisely the kinds issues that need to be addressed before we can entertain the notion of an NTR from the Ground Up. From these basic philosphies, with the engineering realities of an actual design, can come actual flight rules. And right now, that's a pretty tall order.

-- Edited by GoogleNaut at 00:41, 2008-04-05

'GoogleNaut' has got it: the no-flight-at-any-time rule.

I envision an NTR core as something like a spherical colander. The kitchen version sometimes is spherically curved, but when it is, from centre to rim it only covers about 30°. The NTR version would be the rest of the sphere the kitchen one was cut from. Propellant would go in through the little holes, then out the big one on its way to the nozzle constriction.

Earlier he said, "The trouble with an unenriched core is that the core must then be moderated--which adds weight and bulk to the core". What he's missing is the part where, if the propellant is helium, it is also the moderator. Not only that, as it approaches the colander, it's a cold moderator. As such, it produces cold neutrons, which are more strongly absorbed by the fuel. Once inside it is, of course, hot, but that means its density is much reduced. So most of the moderation is done by the cold, dense stuff as it approaches the core.

This provides fission cutoff on turbopump failure, and on turbopump power reduction such that the helium approaching the colander no longer is dense, and it provides fission rate increase if the propellant is pumped in faster, until the gamma and neutron heating of the approaching coolant warms it up, reducing its moderating effectiveness by raising the temperature of the returning neutrons, and restores equilibrium. The fission rate can be controlled entirely by propellant flow.

"A pretty tall order" -- yes, making a highly enriched core that operates near 3,000 K, with a very high number of megawatts per litre, making all that immune to messy failure is indeed that. Impossible, in fact.

Let us look at a nuclear thermal helium rocket with 50 MN liftoff thrust, specific impulse 6400 N·s/kg. Supposing its efficiency in converting heat to the propellant stream's kinetic energy is 0.6, it will need 267 thermal GW. This is about 0.7 as much power as a hydrogen rocket would need, since hydrogen gives more specific impulse. However, that 0.7 power must fit into a vehicle whose nonpropellant mass is 997 tonnes -- out of 4,430 -- rather than 1542 tonnes with hydrogen, because hydrogen gives a lower mass-ratio. Therefore the power/dry mass actually is about eight percent higher.

It has 9.5 km/s of delta 'V' per tankful, and a liftoff mass of 4,430 tonnes, liftoff acceleration 11.278 m/s^2. So its 9.5 km/s, it must go through in 842.37 seconds. Heat production in that time, 224.633 TJ, 62398 MWh.

If the core includes 100 tonnes of uranium, the burnup is 0.624 MWh/kgU, 263 times less than the Darlington plant near me once claimed.

So two days after reaching orbit, or almost reaching orbit and then dispersing, its radioactivity is about the same as would have been that of one of the Darlington reactors, two days after its first startup, if it had run at full power the first day and been shut down throughout the second.

Now suppose the same 62398 thermal MWh came from a 10-tonne highly enriched core. Its dispersal would disperse, of course, the same amount of induced radioactivity, but it would also disperse readily bomb-adaptable materials; possibly in not very many pieces, not very small.

Compared to the enriched material, the radioactivity is a small deal. Sudden loss of coolant at full power guarantees that the core will disperse. The closest we can get to making this dispersal a safe failure is if it disperses only radioactivity, not highly enriched nuclear fuel. As long as there is the possibility of dispersing the latter, even from a never-lit core that has been fumbled by a chemical first stage, the no-actual-flights-at-any-time rule will remain in force.

How shall driving gain nuclear cachet?

G R L Cowan · Veteran Member

I see I worked out an incorrect square-pulse burn time. Dividing the delta 'V' by the initial acceleration is wrong because acceleration would increase. The correct formula starts with thrust over specific impulse, newtons over newton-seconds-per-kilogram yielding kilograms-per-second, the dimensions of the propellant use rate. 50 MN over 6400 N·s/kg gives 7812.5 kg/s.

Mass ratio ~4.4122, initial acceleration 1.15 times 9.80665 m/s^2, therefore initial mass ~4433549 kg, non-propellant mass ~1004837 kg, therefore propellant mass ~3428712 kg, divided by 7812.5 kg/s that's ~438.875 seconds.

This makes the thermal megawatt-hour count 32500 not 62398.

http://www.eagle.ca/~gcowan/boron_blast.html --
let the baby light matches in the fuel storage room!

G R L Cowan · Veteran Member

How about a 1-year progress report?

--- G.R.L. Cowan
Internal combustion made continent:
http://www.eagle.ca/~gcowan/

GoogleNaut · Guru

I am working on what would be needed for a a 24000 MWt fast spectrum or epithermal reactor core for an engine with approximately 1 million lbf with hydrogen...

Trying to get long engine life (long accumulated burn times) with high power is proving a formidable task--as is keeping the specific power density high...

I am still not entirely sure what the geometry of the core should be, although I do lean toward prismatic core elements. I originally was thinking about using solid tungsten with tungsten-uranium alloy fuel inserts...but the self shielding of the fuel will make such a design pretty twitchy...it might be difficult to turn on and turn off...and also prone to spiking...I don't know yet...

Currently I am trying to figure out how to get the Ecole Polytechnique's Dragon 4.0 working. This is an open source, publicly available, multidimensional, multigroup neutron diffusion/transport simulation code that has had very good success in simulating nuclear power cores for many plants including the CANDU. I think that this code can help me decide which geometry would be appropriate,

But I am trying to learn how to use UNIX application on a CYGWIN partition on my harddrive...so I can correctly install and configure Dragon 4.0.

I realize that I am trying to take on a project that would normally keep a moderately sized, staffed and funded research center pretty busy for year or so, so progress is kind of slow... :)

-- Edited by GoogleNaut on Monday 6th of April 2009 06:36:53 PM

Philipum · Veteran Member

GoogleNaut wrote:

I realize that I am trying to take on a project that would normally keep a moderately sized, staffed and funded research center pretty busy for year or so, so progress is kind of slow... :)

-- Edited by GoogleNaut on Monday 6th of April 2009 06:36:53 PM

The NTR from ground to space is one of the most appealing concept to the dreamer I am... Keep working on it, sometimes real progress is made just because some very stubborn individuals actually explore new ideas (it so often happens that researchers proposing original ideas are denied fundings ).

It is a pity that nuclear-based technological challenges cannot be proposed as X-Prize, because you would have a fair chance biggrin

-- Edited by Philipum on Tuesday 14th of April 2009 03:12:41 PM

GoogleNaut · Guru

Thanks Philipum for the kind words...I am starting the design from the standpoint of a reactor operating on a fast fission spectrum, or epithermal spectrum (I haven't decided which is better yet) with initial design parameters of 960 sec of specific impulse in vacuum with liquid hydrogen alone, generating a nominal vacuum thrust of nearly 4.3 MN (960,000 lbf,) operating with a peak core temperature very close to 3000K. Reactor power will approach 24,000 MWt. The engine is also augmented with liquid oxygen, injected as a hot gas downstream of the nozzle throat, to generate a total of about 14.7 MN (3.3 million lbf) in vacuum with an O/F ratio of 4:1, with a combined system Isp of about 650 sec in vacuum with oxygen augmentation. Regernatively cooling the nozzle to an expansion ratio of about 60:1 will be adequate for a ground start for an SSTO type vehicle...attaching a radiation cooled carbon fiber/columbium nozzle extension to an area ratio of 500:1 will give the nearly full 14.7 MN of vacuum thrust. The nozzle exit at the end of radiation cooled extension will be just about 10.2 m in diameter (33.5 ft) which is almost exactly the diameter of the Saturn 5 booster from Project Apollo. This is a big engine! Infact total thermal power (nuclear + combustion) for the engine at full thrust with oxygen augmentation will be just about 54,000 MWt (72 million horsepower.) This would be the most powerful 'steady state' rocket engine ever conceived (only the nuclear pulse Orion [thousands of GW] or Mini Mag Orion [200 GW] are more powerful.) And this engine would--if erosion is completely solved--be the only one that could make the trip from sea level to space without contaminating the launch site and surrounding areas with radioactive fallout.

This engine will be the design basis for a vehicle that will fly from low earth orbit using a "Pod" of three such engines to travel to a comet to mine water ice and bring it back to low earth orbit (or high earth orbit.)

So far, I've bitten off way more than I can chew. But I do have some initial concepts that could be the basis for a design, which I call MINERVA. MINERVA is an acronym for Modular Integrated Nuclear powered Expandible Reusable Vehicle Architecture. I am also looking at not just this vehicle, but a whole new class of launch vechicles that utilize vechicle design philosophies never tried before; a whole new LEO (low earth orbit) and HEO (high earth orbit) space station architicture and associated infrastructure; industrial design and process in zero-g, etc. etc.

It is incredibly ambitious, but I feel confident that it will payoff.

G R L Cowan · Veteran Member

How are you getting on with Dragon 4.0?

--- G.R.L. Cowan ('How fire can be domesticated')
http://www.eagle.ca/~gcowan/

GoogleNaut · Guru

Dragon 4.0 is a UNIX based program...since I don't have unix I am trying to use Cygwin which is a UNIX interpretor for Windows. I suppose Dragon 4.0 could be made to work with a Linux partition, but I have little experience with it, and no experience at all with UNIX.

Once I get the program to work, then I have to learn how to use the modelling language and then use that to create some simple models and analyze them to see if I understand that.

Once I am in a position to actually start working on a real model, I will have to decide what parameters to set it up as: Dragon 4.0 appears able to do multispectral, neutron diffusion calculations, so I would have to make some guesses as to the thermohydraulic geometries...

In short, I've got a hell of a lot of work to do before I can even start. I was told by someone who is an expert (quite a while ago) that what I was trying to do is usually done by a team of doctoral and postdoctoral students, so if I am a little lost, I have good reason to be!

I have some very basic parameters for this reactor that I consider only a jumping off point for an actual conceptual design...

---among other things I do not have a really good idea for allowable fissile burnup for this type of reactor. Allowable burnup would be a function of neutron spectrum, and the fissile loading of the reactor. Allowable burnup will definately be a an indicator of total reactorfull power life. Nozzle throat erosion and neutron embrittlement of the thrust chamber are also very important factors. Once I have an idea of what the reactor can do and have specific nozzle and chamber geometries to work with, I can then turn to computational fluid dynamics to do fun things like compute acoustic resonance modes to eliminate rocket engine problems like whistling. Then we can address vehicle/engine resonance dynamics like 'chugging' and pogo-oscillations.

Whew! I've got my work cut out for me!

Ty Moore

-- Edited by GoogleNaut on Monday 7th of June 2010 03:16:37 AM