Solid core NTR with thermoelectric cooling?

karov · 2005-09-15 06:32:14

http://en.wikipedia.org/wiki/Peltier-Seebeck_effect
http://en.wikipedia.org/wiki/Thermoelectric_cooling

The SCNTR potential is killed by the insufficienlty high melting temperature of the core materials, yes. The Peltier effect is described as depending on the molecule structure of the conductors, their superconductivity parameters, etc... i.e. regarding solid, most often metalic couples in the electric circuit...

but, plasma flow is electric currant. Imagine system where the difference in the electric potential, the voltage builds up between one solid part -- the reactor core and the other fluid - plasma flux. The electric currant is produced and results in cooling the core to be within thee range of the solid state, "draining" the heat and energy in highly controlable and manageable form of electric currant which to turn the reaction mass ( H2O? ) in very hot plasma, channeled by magfield-nozzle in the direction...

What do you think, but reply SERIOUSLY ( !!! ), about the feasibility of such concept?

GCNRevenger · 2005-09-15 11:28:51

You misunderstand the problem you are trying to solve I think: the efficiency of NTR engines depends on the temperature of the propellant. Heat energy always transfers from a hot source to a cold source via thermal conduction, right?

The problem with the core melting is not simply because it gets too hot, the problem is that the core must be at least as hot as the desired propellant temperature, reguardless of any cooling tricks. If your reactor isn't as hot as the propellant temperature you want to acheive, then it the reactor cannot make the propellant any hotter.

If you were to cool the core, by whatever means, all you would accomplish would be to radically LOWER the efficiency because the core can no longer heat the propellant to as high of a temperature. The only practical way to improve the efficiency of a sizeable NTR engines is to increase the melting temperature of the core, since cooling it will do the exact opposit.

karov · 2005-09-16 06:08:37

yes, but transfering heat to the propelant, the heat exchange is actually cooling the core... Imagine such heat transfer from the core, to the plasmed propelant ( mediated perfectly via electric currant in adapted for plasma Peltier transformer ), so effective that the plasma torch receives the reacvtor heat so fast and so efficiently, that the core manages to remain solid -- although producing as much heat that not only to boil it but to evaporate it...

Basicly you can not let heat go from warmer to colder medium without to cool the hotter, isn`t it -- de-misunderstand yourself!

say, Plutonium core in order to possess the necessary electrical characteristics, perhaps should be kept in cryo-conditions ( liquid gases..?)
. The nuclear fission within it will produce heat, but the electric currant -- the propelant plasma itself, will drain it perfectly along the SC currant...

The trick is the heat-generating core and the heat-sucking plasma plume, to be points in Peltier-Seebek chain.

Please, think deeper. I know that in the room temperature solid state physics` list there isn`t material with these features, but in the cold depths of the cryology and the quantum models certainly could be found s.t.

GCNRevenger · 2005-09-16 08:49:24

A clever idea, but I don't think that it will ever be practical... First off, the Peltier effect just can't transfer the billions or tens of billions of watts of energy like that, it just can't. Plasmas nor any known material except at super-low temperatures are superconducting, and superlow temperatures are not achieveable with super-hot plasma nearby. Normal metalic conductors also lose their conductivity the hotter they get, which makes them hotter, which makes them less conductive some more... its a negative feedback loop. And, there is another unwanted feedback loop that you might not have thought of:

-The propellant must be in a plasma state to conduct electricity, regular liquid or gasseous Hydrogen doesn't conduct well at all. The Hydrogen must be heated to an extreme temperature to convert all of it to a plasma.

-Solid core reactors cannot resist temperatures high enough to achieve nearly complete conversion of Hydrogen gas to a plasma.

-The plasma must be in physical contact with the solid Peltier elements of the core in order to conduct electricity and hence cool the core.

Hence the hot plasma, which will be absorbing the heat from the core, will be in physical contact with the core assembly. The heat would travel backwards and melt the core via thermal conduction... Which means that the Peltier effect has even more heat to deal with, both from the nuclear fuel AND from the plasma it is producing, which will get worse and worse the hotter you make the plasma.

In fact, I really doubt that the Peltier effect coupled with a superconductor can transfer much heat at all... If it were that good, it would have already rearranged the war energy is produced and used on Earth. No form of cooling of any kind other then direct thermal transfer to a high-cryogen can cool a high-power nuclear reactor. Nothing.

Austin Stanley · 2005-09-16 13:07:02

A GCNR is going to be to hot for any conventional form of cooling. Their operating temperatures (5000 - 20000 K) are well beyond the melting point for metals and even normaly heat-resistant silcates. So structual material simply cannot be alowed to come in contact with the working medium of a GCNR, it must be contained either in a magnetic field or by a constantly replenished film of super chilled high specific heat substance which is constanly cycled (liquid helium or hydrogen). Realisticly at these temperatures there is realy only one practical cooling option, dumping lots of liquid hydrogen into the chamber which is heated and exhausted. Thankfully this is exactly what a GCNR does

GCNRevenger · 2005-09-16 13:25:22

Bingo Stanley, but Karov here is talking about an exotic solid-core concept, where the reactor dumps its heat into exhaust plasma not via thermal conduction, but instead by electrical conduction like a Peltier cooler. It sounds clever, but I don't think it could possibly work... Thinking more about the thermodynamics of it, a signifigant amount of electricty would be needed to make the heat "move" that I am not sure is accounted for.

As far as a GCNR engine, there is a third route, the one that hypothetical engines actually use: the reaction chaimber is specialy shaped, probobly a toridal configuration, with Hydrogen forming a three-dimensional vortex. In such a vortex, denser materials will settle to the center, and since there is such a huge difference in density between Uranium and liquid Hydrogen, the Uranium will be naturally confined to the center, and not require support by anything but the LH2 flow. Magnets might also be employed to further confine and boost the reaction rate...

A GCNR engine would be hard to build, but it would offer the kind of performance needed to change the rules of the game, and make colonizing Mars a possibility. I doubt a GCNR engine would be any harder to build then this hypothetical themoelectric heated NTR engine karov likes.

karov · 2005-09-17 01:32:43

-The propellant must be in a plasma state to conduct electricity, regular liquid or gasseous Hydrogen doesn't conduct well at all. The Hydrogen must be heated to an extreme temperature to convert all of it to a plasma.
-Solid core reactors cannot resist temperatures high enough to achieve nearly complete conversion of Hydrogen gas to a plasma.
-The plasma must be in physical contact with the solid Peltier elements of the core in order to conduct electricity and hence cool the core.
Hence the hot plasma, which will be absorbing the heat from the core, will be in physical contact with the core assembly. The heat would travel backwards and melt the core via thermal conduction... Which means that the Peltier effect has even more heat to deal with, both from the nuclear fuel AND from the plasma it is producing, which will get worse and worse the hotter you make the plasma.
In fact, I really doubt that the Peltier effect coupled with a superconductor can transfer much heat at all... If it were that good, it would have already rearranged the war energy is produced and used on Earth. No form of cooling of any kind other then direct thermal transfer to a high-cryogen can cool a high-power nuclear reactor. Nothing.

To your points:

1. It is not necessary the whole bulk of hydrogen propelant to be turned in plasma at once... you could begin with several ionized atoms and the process to build up further with the increase of the currant...

2. The reactor will not go warmer cause the heat will be completely pumped out "in real time" via electrical effect. The core is in cryo-conditions, the plasma has not direct heat-transfering contact with the core. The only physical connection between the core and the propelant is electrical currant, passing on SC and further as plasma flux. The plasma flux doesn`t heats the SC cause it is magnetically confined with the SC curant, the SC cryo-system communicates only electrically, the back leak of heat from plasma to core is prevented with thermoinsolation, which is EM field transperant.

3. The plasma ( peltier element two) MUST NOT be in physical contact with the solid Core ( peltier element one) in order to drain heat... induction, magnetic confinement...

GCNRevenger · 2005-09-17 07:36:48

You are trying to have it both ways, but you can't have both:

The Hydrogen plasma IS the other half of the Peltier circut, right? Then in order for there to be a complete circut, then there must be physical contact between either the core or a conductor physically in contact with the core, right?

No material of any kind can remain solid if it is in contact with superhigh temperature plasmas, which will include this "passing on SC" thing. If the plasma won't melt the core, it will melt the superconductor bridge just the same. Plus, since the plasma is so hot, it can't be a superconductor. To top it off, the superconductor bridge to the core cannot itself be cooled by the Peltier effect, and since it is contact with the plasma, there is no way to keep it solid.

Anyway, this is all irrelivent, because there is no possible way that you could move billions or tens of billions of watts of heat with the Peltier effect. I am not even convinced that your idea would move any heat at all, since the Peltier effect requires an input of electricity to "move" heat, which makes sense to me thermodynamically since you are effecting a decrease in entropy in the core.

It is just easier to build a GCNR engine, the vortex chaimber concept has been validated, and the Russians were even thinking about building one in the late NTR heyday.

ftlwright · 2005-09-20 13:26:54

Honestly if you have that much control over the plasma your better off with a MHD derived design. Perhaps a hybrid between quasi-symmetrical stellarator http://www.pppl.gov/projects/pages/ncsx.html where highly energetic plasma is bled off and feed into a more traditional Electric Propulsion rocket or VASIMR type engine. The point of of going solid or gas core is so that you don't have to deal with the *cough* elegance *cough* of control plasmas. If you have that type of control, you really don't need either of these solutions.

GCNRevenger · 2005-09-20 16:53:26

That thing looks like a fusion reactor... we're assuming here that the power source is still plain old thermal fission. The VASIMR engine is the engine of choice to use electricity to make plasma... Although if you did have a fusion power plant, using that to feed a VASIMR engine without the heating chaimber would be swell.

karov · 2005-09-24 01:04:35

You are trying to have it both ways, but you can't have both:
The Hydrogen plasma IS the other half of the Peltier circut, right? Then in order for there to be a complete circut, then there must be physical contact between either the core or a conductor physically in contact with the core, right?
No material of any kind can remain solid if it is in contact with superhigh temperature plasmas, which will include this "passing on SC" thing. If the plasma won't melt the core, it will melt the superconductor bridge just the same. Plus, since the plasma is so hot, it can't be a superconductor. To top it off, the superconductor bridge to the core cannot itself be cooled by the Peltier effect, and since it is contact with the plasma, there is no way to keep it solid.
Anyway, this is all irrelivent, because there is no possible way that you could move billions or tens of billions of watts of heat with the Peltier effect. I am not even convinced that your idea would move any heat at all, since the Peltier effect requires an input of electricity to "move" heat, which makes sense to me thermodynamically since you are effecting a decrease in entropy in the core.
It is just easier to build a GCNR engine, the vortex chaimber concept has been validated, and the Russians were even thinking about building one in the late NTR heyday.

1. The contact between the plasma and the core ( the two Peltier points of the circuit ) is only electromagnetical ( induction) -- no thermal conducting between them.

2. Yes, the Peltier effect needs electric currant to run, but what about feeding in the electricity via Seebek effect -- both effects coupled in a positive backloop. Don`t tell me that this brakes the thermodynamical laws -- there is gradient of entropy -- the spiting of superhot plasma is due to degradation of the fissible material in the core.

At the temperature of the liquid hydrogen the whole core if immersed in the fuel tank of LOH will be superconductive...

In order to visualise it better lets regard the scheme as 2-in-1 of solenoid+fission reactor.

Please tame your nerves, and understand that I DO NOT insist that this is technologically feasible, but just that it is usefull to look at other tech alternatives, hidding very offten, unexpectedly cheap and easy solutions of many multidecade problems, yeah?

GCNRevenger · 2005-09-24 01:36:34

Nonsense, you can't have a single device operate on both a Peltier and Seebek effect, since the two effects are the opposite of one another. It is not thermodynamically possible for a single circut to have both effects.

"is only electromagnetical ( induction) -- no thermal conducting between them"

This is not possible. Either the plasma is itself in physical contact with the reactor fuel elements, in which case the superhigh temperatures will melt the core, or else the plasma is in physical contact with a conductive bridge (probobly superconducting wire), in which case it will melt that too.

In either event, you can't move enough of the heat produced by the core to the plasma via the Peltier effect, there is just too much of it. It would require so much electricity that the Peltier generator would be bigger then the whole engine I would imagine.

You aren't getting anywhere with this...

No, I don't think that it is useful to look for clever physics tricks, ways of cheating, or other "hidden technologies" because they are basically without exception full of bunk. I feel pretty strongly that beliefe in physics tricks is foolish nonsense, that extremely simple but difficult physics problems, like thermal rocketry, are never overcome by "cheap and clever" ways of cheating.

karov · 2005-09-24 02:34:16

Very-very emotional reply Revenger!

1. What about TWO circuits combined - one for Seebek, one for the Peltier ( the heat increase or attempt hor increase creates currant, the currant drains the heat )? Possitive feedaback.

2. Why do you think that it is not possible to interact with plasma without physical ( as I understand you mean the plasma to hit directly the electrode) contact?
Imagine plasma contained in electromagnetic field created by the electric currant generated by the mutually amplifying eachother Peltier and Seebeck effects. The changes of the electromagnetic field will be transfered to the plasma, the changes in the plasma to the cryo-solenoid reactor... this is electric circuit too.

GCNRevenger · 2005-09-24 08:41:33

"one for Seebek, one for the Peltier ( the heat increase or attempt hor increase creates currant, the currant drains the heat )? Possitive feedaback."

No, this won't work, because you will need more electricity to cause the temperature difference via Peltier then this same temperature difference can produce via Seebek. Simple thermodynamics, you can't get something for nothing. Given the inefficiency of the Peltier effect, the excess heat would be more then enough to ruin the reactor I would think.

So you want to attach the supercondutor bridge to a microwave generator to avoid having to touch the plasma? No no, this is not an electric circut, it involves two conversions, heat to electricity to microwave. This won't work, because the whole Peltier effect relies upon the potential difference of the heat source and the heat sink. By adding the microwave conversion step, the two halves of the Peltier circut are no longer connected, and hence will not work.

The microwave generator itself cannot convert electricity into microwave radiation without itself producing heat; if you try to pump billions of watts of electricty into a reasonable generator, it will melt too.

New Mars Forums

Announcement

#1 2005-09-15 06:32:14

Re: Solid core NTR with thermoelectric cooling?

#2 2005-09-15 11:28:51

Re: Solid core NTR with thermoelectric cooling?

#3 2005-09-16 06:08:37

Re: Solid core NTR with thermoelectric cooling?

#4 2005-09-16 08:49:24

Re: Solid core NTR with thermoelectric cooling?

#5 2005-09-16 13:07:02

Re: Solid core NTR with thermoelectric cooling?

#6 2005-09-16 13:25:22

Re: Solid core NTR with thermoelectric cooling?

#7 2005-09-17 01:32:43

Re: Solid core NTR with thermoelectric cooling?

#8 2005-09-17 07:36:48

Re: Solid core NTR with thermoelectric cooling?

#9 2005-09-20 13:26:54

Re: Solid core NTR with thermoelectric cooling?

#10 2005-09-20 16:53:26

Re: Solid core NTR with thermoelectric cooling?

#11 2005-09-24 01:04:35

Re: Solid core NTR with thermoelectric cooling?

#12 2005-09-24 01:36:34

Re: Solid core NTR with thermoelectric cooling?

#13 2005-09-24 02:34:16

Re: Solid core NTR with thermoelectric cooling?

#14 2005-09-24 08:41:33

Re: Solid core NTR with thermoelectric cooling?

Board footer