Author Topic: Thermoelectrically Harvesting Waste Heat from Rocket Combustion  (Read 1689 times)

Offline sanman

  • Senior Member
  • *****
  • Posts: 4000
  • Liked: 550
  • Likes Given: 7
I was reading about newer discoveries in "topological" or Dirac materials which can lead to more efficient thermoelectric devices:

http://news.mit.edu/2018/topological-materials-turning-heat-electricity-0117

http://www.pnas.org/content/early/2018/01/09/1715477115.abstract

Recently, Rocketlab's Electron rocket has been in the headlines for delivering small payloads to orbit. That rocket is notable for its use of a battery-powered electric turbopump, which reduces the plumbing complexity for the engine, while also improving reliability. However, a key drawback is that batteries obviously have a much inferior energy density compared to chemical fuel, resulting in a mass penalty for the rocket. (The Electron rocket even drops used-up batteries during ascent in order to reduce parasitic weight.)

Could thermoelectrics be used to solve the battery problem, or at least usefully alleviate it?

Given that a chemical propellant mixture is already being used to generate thrust combustively, could some of that waste heat flux from combustion be used to power an electric turbopump, such as in the case of the Electron rocket?

If currently available thermoelectric materials aren't enough to do the job, then what kind of performance improvement would be required?
« Last Edit: 01/22/2018 10:03 PM by sanman »

Offline DarkenedOne

  • Full Member
  • ****
  • Posts: 948
  • Liked: 55
  • Likes Given: 6
You could ask the same question regarding the waste heat from a car or an airplane.  The tradeoff is that you are increasing efficiency at the cost of increasing complexity, cost, and weight.  For a rocket it is hard to have a thermoelectric system to recover the waste heat that would be worth the extra mass and system complexity. 

Although I would like to point out the expander cycle does use the waste heat from the combustion chamber to drive the rocket's turbopump.

Offline sanman

  • Senior Member
  • *****
  • Posts: 4000
  • Liked: 550
  • Likes Given: 7
Okay, but weight isn't as quite as critical for cars, or even airplanes, as it is for rockets.

With a rocket like the Electron using batteries, I'm wondering if efficient thermoelectrics could provide a mass-savings relative to batteries, without necessarily being overly complex.

Plus Rocketlab uses 3D-printing for fabrication of key components - so I'm wondering if thermoelectric sheath could likewise be integrated through the 3D-printing process, to harvest waste heat flux.
« Last Edit: 01/22/2018 10:27 PM by sanman »

Offline Jim

  • Night Gator
  • Senior Member
  • *****
  • Posts: 31825
  • Cape Canaveral Spaceport
  • Liked: 10383
  • Likes Given: 317

Could thermoelectrics be used to solve the battery problem, or at least usefully alleviate it?


No, there is no free lunch.  Still need battery for startup and the thermoelectrics:
a.  Require a radiator
b.  Would reduce the rocket output by pulling heat out.


Just stop this nonsensical search for energy for nothing or unobtainium.

« Last Edit: 01/22/2018 11:00 PM by Jim »

Offline sanman

  • Senior Member
  • *****
  • Posts: 4000
  • Liked: 550
  • Likes Given: 7
Hi Jim,

No, there is no free lunch.  Still need battery for startup

Alright, but battery for startup would be less than battery for full flight, so there could still be useful mass savings.
Also, if startup is on the pad, then maybe battery could be left on the pad.


Quote
and the thermoelectrics:
a.  Require a radiator

Consider that radiative cooling is not your only option when you have onboard cryo-propellant as a cold sink (eg. LOX, LCH3, LH2, etc)
Couldn't thermoelectric harvesting of heat flux be used as part of expander cycle / regenerative cooling / ?


Quote
b.  Would reduce the rocket output by pulling heat out.

Sure, but arguably even expander cycle's heat exchange does that. When it's being done for a useful purpose, like running the turbopump, then it can be an acceptable trade-off. After all, your combustion chamber isn't going to get fed fuel/oxidizer for free.

Quote
Just stop this nonsensical search for energy for nothing or unobtainium.

But it's not supposed to cost nothing - it's supposed to be a trade-off. RocketLab chose the electric turbopump approach for Electron because they didn't want to deal with the plumbing headache (ie. wiring is supposedly easier than plumbing, and more reliable). But that then introduces the mass penalty from using batteries, due to their lower energy density. I want to know if thermoelectrics can reasonably offer weight savings compared to batteries - especially as you scale up the rocket size.
« Last Edit: 01/24/2018 04:55 PM by sanman »

Offline BeyondNERVA

  • Member
  • Posts: 17
  • Liked: 28
  • Likes Given: 9
I think I generally agree with Jim, although I see what you're trying to get at.

If it were me, I'd look at thermionics rather than thermoelectrics. They tend to be significantly more efficient at higher temperatures, and there's been a fair bit of work done on them in recent years coupled to direct solar power production.

I guess the question is "how much of the thermal energy that the active cooling of the nozzle captures is being used later in the combustion cycle?" If you're just adding weight to steal energy from another part of the system, you're shooting yourself in the foot. If it truly is waste energy, and you can get the harvesting system lightweight and efficient enough, that's a different story.

Offline Jim

  • Night Gator
  • Senior Member
  • *****
  • Posts: 31825
  • Cape Canaveral Spaceport
  • Liked: 10383
  • Likes Given: 317
I want to know if thermoelectrics can reasonably offer weight savings compared to batteries - especially as you scale up the rocket size.


They don't.  They are too inefficient.

Offline Jim

  • Night Gator
  • Senior Member
  • *****
  • Posts: 31825
  • Cape Canaveral Spaceport
  • Liked: 10383
  • Likes Given: 317

Consider that radiative cooling is not your only option when you have onboard cryo-propellant as a cold sink (eg. LOX, LCH3, LH2, etc)
Couldn't thermoelectric harvesting of heat flux be used as part of expander cycle / regenerative cooling / ?


No, you don't want to dump it into the propellant

No, because you are taking more energy out of the system.

Offline chapi

  • Member
  • Posts: 43
  • Liked: 5
  • Likes Given: 5

Consider that radiative cooling is not your only option when you have onboard cryo-propellant as a cold sink (eg. LOX, LCH3, LH2, etc)
Couldn't thermoelectric harvesting of heat flux be used as part of expander cycle / regenerative cooling / ?


No, you don't want to dump it into the propellant

No, because you are taking more energy out of the system.

Well, one could argue that on an expander engine, you could use the remaining radiative energy (which is literally lost in space) of nozzle extension.

Some engines for instance do feature a heat shield that protect the upper part of the engine and the cryo equipments of the stage from heat radiated by the nozzle extension. One side hot, the other side close to cryo systems...you might think of making the heat shield a Seebeck effect electric generator.

But doing the math, you end up with not much power. Better add a battery for such a short mission as a launch.


« Last Edit: 01/26/2018 07:55 PM by chapi »

Offline sanman

  • Senior Member
  • *****
  • Posts: 4000
  • Liked: 550
  • Likes Given: 7
I think I generally agree with Jim, although I see what you're trying to get at.

If it were me, I'd look at thermionics rather than thermoelectrics. They tend to be significantly more efficient at higher temperatures, and there's been a fair bit of work done on them in recent years coupled to direct solar power production.

Thanks for that - I'd read about the promise of nano-gap thermionics a long time ago, and forgotten about it.

Here's a recent article:

https://www.sciencedirect.com/science/article/pii/S2211285516305912

Quote
Thermionic energy converters (TECs) are a direct heat-to-electricity conversion technology with great potential for high efficiency and scalability. However, space charge barrier in the inter-electrode gap and high anode work function are major obstacles toward realizing high efficiency. Here, we demonstrate for the first time a prototype TEC using a back-gated graphene anode, a barium dispenser cathode, and a controllable inter-electrode gap as small as 17 m, which simultaneously addresses these two obstacles. This leads to an electronic conversion efficiency of 9.8% at cathode temperature of 1000 C, the highest reported by far. We first demonstrate that electrostatic gating of graphene by a 20 nm HfO2 dielectric layer changes the graphene anode work function by 0.63 eV, as observed from the current-voltage characteristics of the TEC. Next, we show that the efficiency increases by a factor of 30.6 by reducing the gap from 1 mm down to 17 m, after a mono-layer of Ba is deposited on graphene by the dispenser cathode. Finally, we show that electrostatic gating of graphene further reduces the graphene work function from 1.85 to 1.69 eV, leading to an additional 67% enhancement in TEC efficiency. Note that the overall efficiency using the back-gated graphene anode is 6.7 times higher compared with that of a TEC with a tungsten anode and the same inter-electrode gap.


So, using Rocket Lab's Rutherford engine as an example, I'm imagining that its operating temperature is above 1000C, even though it's a 3D-printed engine made by EBM.

So graphene/nanotubes/SP2carbon is notorious for its low work function. Cathode-Ray Tubes made from CNTs have been shown to be very efficient, for example. Graphene has the high conductivity. And the HafniumOxide is famous as an insulator used in microchips - I guess that's what prevents the emitted electrons from crossing back over, and helps maintain the one-way bias. But I'm not sure what the Barium does - although I've read that Barium sometimes gets used as an intercalation agent in various materials applications.
 

I guess the question is "how much of the thermal energy that the active cooling of the nozzle captures is being used later in the combustion cycle?" If you're just adding weight to steal energy from another part of the system, you're shooting yourself in the foot. If it truly is waste energy, and you can get the harvesting system lightweight and efficient enough, that's a different story.


Well, Rocketlab's Electron rocket uses LiPo batteries, whose energy density is much inferior to chemical fuel.

Supposing we use the ~10% thermionic conversion efficiency cited in the article above. Then in order to generate the 1 MW of electrical power that the LiPo batteries generate for Electron's electric turbopump, it means there needs to be ~10 MW of waste heat being coupled with the thermionic layers.

Note that Rocketlab's Rutherford engines are smaller, and Electron relies on clustering them. Clustering smaller engines translates to comparatively more available surface area to use for coupling with the thermionics.
« Last Edit: 01/27/2018 06:02 AM by sanman »

Offline Jim

  • Night Gator
  • Senior Member
  • *****
  • Posts: 31825
  • Cape Canaveral Spaceport
  • Liked: 10383
  • Likes Given: 317

So, using Rocket Lab's Rutherford engine as an example, I'm imagining that its operating temperature is above 1000C, even though it's a 3D-printed engine made by EBM.


doubtful, would be less
Then in order to generate the 1 MW of electrical power that the LiPo batteries generate for Electron's electric turbopump, it means there needs to be ~10 MW of waste heat being coupled with the thermionic layers.


Not going to get 10MW


Note that Rocketlab's Rutherford engines are smaller, and Electron relies on clustering them. Clustering smaller engines translates to comparatively more available surface area to use for coupling with the thermionics.


Not really.  They are too close together.
« Last Edit: 01/27/2018 01:38 PM by Jim »

Offline ZachF

  • Full Member
  • ****
  • Posts: 406
  • NH, USA, Earth
  • Liked: 263
  • Likes Given: 78
This exercise seems a bit silly since the expander cycle does pretty much everything you want this to do simpler and easier.

Offline sanman

  • Senior Member
  • *****
  • Posts: 4000
  • Liked: 550
  • Likes Given: 7
This exercise seems a bit silly since the expander cycle does pretty much everything you want this to do simpler and easier.

Traditional Expander Cycle relies on plumbing for mass transport necessary to effect the heat exchange.

Meanwhile, aren't electrons also useful mediators of thermal energy in their own right, which might potentially be easier to manage?
Electric turbopump can be switched on/off at will, without worrying about plumbing issues, or about temperature or pressure conditions to make it work properly. Electric is more reliable, and easily restartable - which could benefit reusability in the future.
Thermoelectrics/Thermionics is "Solid-State" having no moving parts.
« Last Edit: 01/28/2018 11:31 PM by sanman »

Offline Jim

  • Night Gator
  • Senior Member
  • *****
  • Posts: 31825
  • Cape Canaveral Spaceport
  • Liked: 10383
  • Likes Given: 317
just stop

Offline john smith 19

  • Senior Member
  • *****
  • Posts: 6546
  • Everyplaceelse
  • Liked: 918
  • Likes Given: 5833
Electric turbopump can be switched on/off at will, without worrying about plumbing issues, or about temperature or pressure conditions to make it work properly.
Very dependent on what the propellants are. If you're talking cryogenics they might boil, so the pump (designed to pump liquids) is trying to pump vapour. massively over speeds and rips itself apart.
Quote from: sanman
Electric is more reliable, and easily restartable - which could benefit reusability in the future.
Thermoelectrics/Thermionics is "Solid-State" having no moving parts.
You have a chicken and egg situation.

Start engine --> electric output --> drive pump.

You'd get a faster response to connecting a generator to the drive shaft of the pump.

As the temperature falls the ability for a nozzle to radiate the necessary power falls and the size of the radiators gets a lot bigger.

It's one of those ideas that can work (sort of) but is such a colossal PITA it's not worth it.  :(

I think that's going to be my last post on this topic.
BFS. The worlds first Methane fueled FFORSC engined CFRP structured A380 sized aerospaceplane tail sitter capable of flying in Earth and Mars atmospheres. BFR. The worlds biggest Methane fueled FFORSC engined CFRP structured booster for BFS. First flight to Mars 2022. Forward looking statements. T&C Apply

Offline sanman

  • Senior Member
  • *****
  • Posts: 4000
  • Liked: 550
  • Likes Given: 7
Electric turbopump can be switched on/off at will, without worrying about plumbing issues, or about temperature or pressure conditions to make it work properly.
Very dependent on what the propellants are. If you're talking cryogenics they might boil, so the pump (designed to pump liquids) is trying to pump vapour. massively over speeds and rips itself apart.
Quote from: sanman
Electric is more reliable, and easily restartable - which could benefit reusability in the future.
Thermoelectrics/Thermionics is "Solid-State" having no moving parts.
You have a chicken and egg situation.

Start engine --> electric output --> drive pump.

You'd get a faster response to connecting a generator to the drive shaft of the pump.

So that's like worrying about a turbine spooling up. Initially you could have some starter current coming from an electrical power supply on the launch pad, to "spool up" the electric turbopumpand get it working at full load. Aside from that, some electrical power could be stored in an ultra-capacitor as a buffer, which then minimizes your response time. Fair enough?

 
Quote
As the temperature falls the ability for a nozzle to radiate the necessary power falls and the size of the radiators gets a lot bigger.

But the temperature from rocket combustion should be quite high, and within the range of thermionic conversion. Rocket Lab's Rutherford engine may be made of lower-temperature materials because it's EBM-manufactured, but it's got regenerative cooling which allows it to operate at higher temperatures than would otherwise be possible.

Quote
It's one of those ideas that can work (sort of) but is such a colossal PITA it's not worth it.  :(

But why see the benefits as only marginal, when it would save on battery mass (which weighs quite a lot)?
Do you feel it would be like a Rube-Goldberg machine and unnecessarily complex?

I disagree with the argument that a large radiator would necessarily be required, when absolute temperature is already high, and when cryopropellant could also be involved for the cold side. And it wouldn't have to be a full-blown expander cycle either, since you only have to maintain enough temperature difference sufficient to meet your thermionic electric output requirements (ie. the plumbing could be simpler than what's needed for traditional expander cycle)
« Last Edit: 02/04/2018 05:08 AM by sanman »

Online A_M_Swallow

  • Elite Veteran
  • Senior Member
  • *****
  • Posts: 8225
  • South coast of England
  • Liked: 269
  • Likes Given: 109
The first stage of a launch vehicle is only used for a few minutes so the energy to be recycled is small. SEP tugs use their main engine for months, see if their waste energy can be reused. The human compartments on space craft and space ships are at about 300 K and need cooling, particularly in direct sunlight. Some of that waste heat could be recycled via the batteries.

Offline IRobot

  • Full Member
  • ****
  • Posts: 1204
  • Portugal & Germany
  • Liked: 222
  • Likes Given: 208


No, there is no free lunch.  Still need battery for startup

Alright, but battery for startup would be less than battery for full flight, so there could still be useful mass savings.
Also, if startup is on the pad, then maybe battery could be left on the pad.

If there is something that SpaceX and Rocketlab show vs "old space" is that simplicity is the key to lower price, not increasing performance.
And you are just telling them to go the opposite way.

Offline Jim

  • Night Gator
  • Senior Member
  • *****
  • Posts: 31825
  • Cape Canaveral Spaceport
  • Liked: 10383
  • Likes Given: 317

But why see the benefits as only marginal, when it would save on battery mass (which weighs quite a lot)?
Do you feel it would be like a Rube-Goldberg machine and unnecessarily complex?


The battery doesn't weigh that much and the added complexity is not worth the effort.  It would cost more.  Cost is more important than absolute performance.  If they wanted more mass to orbit and better performance, they would be using a standard engine with a turbopump and not one with batteries.

Rube-Goldberg because there is multiple systems doing the same thing.  You are trying to optimize the wrong thing.

Tags: