Author Topic: Advanced Radiators for MMW NEP  (Read 2683 times)

Offline rusty

  • Full Member
  • *
  • Posts: 191
  • Liked: 14
  • Likes Given: 18
Advanced Radiators for MMW NEP
« on: 01/26/2012 10:09 am »
A discussion of existing or new concepts and any ideas for low mass radiators for multi-megawatt (MMW) nuclear electric propulsion (NEP).

To start off - One idea that's been around for over 40yrs is the liquid droplet radiator. There are several variants of this theme, reviewable in this 1989 NASA summary.
http://ntrs.nasa.gov/archive/nasa/casi.ntrs.nasa.gov/19900002489_1990002489.pdf
A brief overview of the basic droplet design using charged liquid drawn across a high tempurature fabric also available in this 2007 document.
http://courses.ucsd.edu/rherz/mae221a/reports/Nelson_221A_F07.pdf

Another variant is the rotating disc liquid droplet radiator. I'd appreciate any clarification as this appears to replace charging the liquid with centrifugal force to carry it from droplet generator to collector.

I'd propose yet another variant; a combination of the rotating sheet radiator, but using droplets. This would effectively mimic a chocolate chip belt, where the droplets do not traverse a surface as in other designs, instead the surface (fabric) carries them to a collector.

Offline rusty

  • Full Member
  • *
  • Posts: 191
  • Liked: 14
  • Likes Given: 18
Re: Advanced Radiators for MMW NEP
« Reply #1 on: 01/26/2012 10:15 am »
A more near-term proposal for low mass radiators are fabric heat pipes. I've seen a few theories calling for carbon fiber or ceramic fabric. These operate as typical heat pipe/fin radiators except the material is light weight fabric instead of metallic.

The clear advantage appears to be weight and launch packaging, but I haven't come across any recent studies. Any links or information on this concept is appreciated.
---------------------------------------------------------
*edit:
I'm often dubious of material science approaches to problems, as it becomes a matter of benefits vs consequence. In this case, any reduction in conductivity or irradiance from carbon or ceramic radiators results in an increase in the coolant required.

This materials approach is in contrast to a fundamental (design) approach that can often make significant gains. ie; I've read reports of aforementioned advanced material heat pipes weighing 10% less than metallic systems, while liquid metal droplet systems would weigh half as much. For a spacecraft that would typically budget 40% of its mass toward heat rejection, that difference is quite significant and possibly why I've seen very little regarding advanced material radiators.
« Last Edit: 01/27/2012 08:28 am by rusty »

Offline 93143

  • Senior Member
  • *****
  • Posts: 3038
  • Liked: 294
  • Likes Given: 1
Re: Advanced Radiators for MMW NEP
« Reply #2 on: 01/26/2012 11:33 am »
Keep in mind that if you're shedding heat from a reactor, you want it at high temperature.  Power radiated per unit area is proportional to the fourth power of temperature; ie: at 1200 K it's 16 times as high as at 600 K, or 256 times as high as at 300 K.

For a perfect Carnot cycle, the minimum radiator area is achieved at a thermal efficiency of 25%, meaning that the radiators are at 3/4 of the reactor core temperature...

Offline rusty

  • Full Member
  • *
  • Posts: 191
  • Liked: 14
  • Likes Given: 18
Re: Advanced Radiators for MMW NEP
« Reply #3 on: 01/29/2012 04:38 pm »
...
For a perfect Carnot cycle, the minimum radiator area is achieved at a thermal efficiency of 25%, meaning that the radiators are at 3/4 of the reactor core temperature.
An interesting tidbit, but one that does not apply to liquid droplet or other advanced heat rejection systems.

Note that I've said liquid droplet, not liquid metal droplet. The reason being metal need not be used in a liquid droplet design. A spinning disk configuration eliminates the electromagnetic necessity while a belt configuration reduces surface tension requirements. This allows the use of a number of different molecules as the secondary working fluid.

Most appealing, IMO, of a belt configuration is the further possibility of phase transition (ie; The droplets becoming solid prior to collection as in the "chocolate chip" comparison.) A secondary heat exchanger would melt the droplets/chips before the fluid is passed through the primary heat exchanger on way to the belt/droplet radiator, significantly improving thermal transfer from the primary to secondary working fluids.

Offline 93143

  • Senior Member
  • *****
  • Posts: 3038
  • Liked: 294
  • Likes Given: 1
Re: Advanced Radiators for MMW NEP
« Reply #4 on: 01/29/2012 06:50 pm »
Care to explain how such a system would sidestep the laws of thermodynamics?  That derivation is pretty fundamental.

Offline rusty

  • Full Member
  • *
  • Posts: 191
  • Liked: 14
  • Likes Given: 18
Re: Advanced Radiators for MMW NEP
« Reply #5 on: 02/05/2012 07:12 am »
A bit surprised by the lack of interest, or even corrections, but I'll offer my planned elaborations anyways to stir thought...

1) In the second post I mentioned fabric radiators, even though most advanced material approaches use solid composites. The stowage comment previously mentioned is a definite advantage of 'soft' radiators and there's a strong possibility of weight reduction and greater efficiency as well. Of course they would have to be 'strung' in some manner to keep them in place in advent of pressure loss, but overall it appears to be a natural progression in advanced composite design.
Specifically; A laminate or layered 'soft' radiator could use high strength fabric (carbon fiber) as the structural element, a metallic inner film to maintain pressure and increase conductivity, the exterior could be impregnated with ceramic or metal to improve irradiance.
But with material science, as mentioned, there's consequences to go with the advantages. An 'inflated' radiator will have more volume per surface area while the surface's 'weave' inherently creates depth and greater surface.

2) Another configuration of the liquid droplet design I listed above could do away with the secondary generating capability (the secondary working fluid then becomes coolant). As such, the phase transition (melting) heat exchanger would then be moved from after the primary heat exchanger to before.
While only analysis of complete systems can determine which is better, I'd veer toward this design. The sole working fluid then becomes 'flash-cooled', improving pressure differential and thus generating capacity, but at the expense of a dual-stage system. The coolant would also head to the droplet radiator at a lower temp. The main advantage is a simpler and possibly lighter system that's nearly as efficient.

Offline A_M_Swallow

  • Elite Veteran
  • Senior Member
  • *****
  • Posts: 8679
  • South coast of England
  • Liked: 403
  • Likes Given: 180
Re: Advanced Radiators for MMW NEP
« Reply #6 on: 02/05/2012 02:20 pm »
A bit surprised by the lack of interest, or even corrections, but I'll offer my planned elaborations anyways to stir thought...
{snip}

Do not worry about it, NASA's proposal for small (1kW) nuclear reactors were ignored as well.

The thrusters for solar electric propulsion also need cooling, although the radiators frequently get combined with the solar arrays.

Offline alexterrell

  • Full Member
  • ****
  • Posts: 1658
  • Germany
  • Liked: 102
  • Likes Given: 71
Re: Advanced Radiators for MMW NEP
« Reply #7 on: 02/07/2012 06:17 am »
Keep in mind that if you're shedding heat from a reactor, you want it at high temperature.  Power radiated per unit area is proportional to the fourth power of temperature; ie: at 1200 K it's 16 times as high as at 600 K, or 256 times as high as at 300 K.

For a perfect Carnot cycle, the minimum radiator area is achieved at a thermal efficiency of 25%, meaning that the radiators are at 3/4 of the reactor core temperature...
So the biggest benefit for NEP is to work on high temperature reactors - of the sort proposed on Earth for chemical synthesis of Hydrogen - at about 1500K outlet temperature.

On Earth you'd then have a 320K "radiator" and 50% electric efficiency, but in space you might go to 800K for the radiator and accept 30-35% efficiency - so not as extreme as 25% and 3/4.

Pebble bed modular reactors might do, though they have a low power density core. Kirk Sorrenson would know what reactor chemistry works well with these bounds.
« Last Edit: 02/07/2012 06:20 am by alexterrell »

Offline 93143

  • Senior Member
  • *****
  • Posts: 3038
  • Liked: 294
  • Likes Given: 1
Re: Advanced Radiators for MMW NEP
« Reply #8 on: 02/07/2012 06:56 am »
The radiator temperature corresponding to minimum area increases from 75% of the hot-side temperature as the real system efficiency declines from the Carnot value.  On the other hand, the radiator mass is not the only mass in the system, so the actual optimum will not be the minimum-area value.  The lighter the radiators, the more the conclusion will be affected by the rest of the system.  Still, the Stefan-Boltzmann law is very steep...

Offline alexterrell

  • Full Member
  • ****
  • Posts: 1658
  • Germany
  • Liked: 102
  • Likes Given: 71
Re: Advanced Radiators for MMW NEP
« Reply #9 on: 02/07/2012 01:26 pm »
Very steep indeed, but once you get to 600K or so you've got a pretty effective radiator, so reducing this further at the expense of upping your reactor size may not make much sense.

If you get some real world numbers you can optimise this quite easily, but what ever happens, the hotter you can make your hot side the better.

Mass = Reactor Mass + Radiator Mass
Mass = K1 x (Th - Tl)/Th + K2 / Tl^4

Where K1 and K2 are quasi constants.


The PBMR ran helium through the nuclear reactor directly into turbines. I suspect the limiting factor is the turbine temperature rather than the nuclear fuel. But using helium is not great for your reactor power density, but means one less heat exchanger.

Tags: