NASASpaceFlight.com Forum
Commercial and US Government Launch Vehicles => Rocket Lab => Topic started by: Asteroza on 04/15/2015 10:42 am
-
So Rocket Labs USA recently revealed their Rutherford engine, which uses electric oxidizer/fuel turbopumps to up their performance.
http://www.rocketlabusa.com/about-us/propulsion/rutherford/ (http://www.rocketlabusa.com/about-us/propulsion/rutherford/)
It sorta decouples turbopump performance from some of the engine conditions, which means you can do some potentially interesting tricks. Perhaps constant injector pressure at various throttle rates?
Anyways, the weird thought I had was, there is another case where constant power was potentially interesting. For an air turbo rocket, using a (solid?) fuel rich gas generator to drive a turbine at constant speed, to drive an air fan/compressor for afterburning the fuel rich exhaust. What about an electric fan ATR rutherford engine? The power consumption would be atrocious though, requiring huge power sources/batteries, unless you do some sick MHD power generator setup on the exhaust like the various JP Aerospace proposals, or commit to a gas generator exhaust turbine power generator upstream of air breathing combustion (resembling a conventional turbojet in cycle but with split shafts) (or a parasitic power turbine running off nozzle regen cooling gas expansion?).
It would decouple a few points on the ATR though, notably one could decrease oxidizer turbine power on the gas generator segment to trade with fan power on the airbreathing segment but not necessarily in a 1:1 ratio, as well as temporarily boosting fuel turbine power during the airbreathing phase.
-
Electrically driven pumps for a rocket would need an impressive electrical supply, which is almost certainly going to be heavier than a gas generator and turbine. But I see a few advantages:
- Rapid dynamic adjustments to propellant delivery, independent for each fluid if desired. Adjusting mix on the go becomes easy.
- With the motor as an element sealed in with the pump, allowing the pumped fluid to be the bearing working fluid, bearing seals can be much easier.
- With a battery supplying power the pumps can start almost instantly without waiting for a turbine to come up to speed.
- A generator going out on one engine of a cluster need not shut down that engine if other engines can spare the electricity.
-
I was also very interested in this news... I figured they were pressure fed. Interesting that this traded well against pressure fed. They both must impact dry mass pretty significantly. For comparison this is pretty close to SpaceX Kestrel thrust and better ISP.
The thought I had was that this might allow a path for low development costs on a kick stage. Currently these are often solids so ISP >300 is doing well. This would allow storable, safe biprops like N2O+Propane.
-
Electrically driven pumps for a rocket would need an impressive electrical supply
A megawatt apparently: https://twitter.com/rocketlabusa/status/588370273583271936
which is almost certainly going to be heavier than a gas generator and turbine
That's what I figured too, but it's been running around my head and I realized battery+motor tends to do quite well on power to weight. Where it's not great is energy density for long duration. But a chemical rocket stage doesn't care about long duration, 3 minutes tops for the first stage. 3 minutes of 1 megawatt is 50 kilowatt hours, top of the line lithium-polymer might be able to do that in <200 kg. And the alternative likely isn't gas generator, it would be pressure fed for this size and simplicity of engine, which has its own mass penalties and pressurized gas is even worse on energy density. The more I think about it, the cleverer I think this is.
The other thing to consider with replacing a number of systems with batteries is that you're replacing multiple complex problems with one simple problem, and buying that off the shelf.
Sure, this probably doesn't completely close the gap to a good gas generator engine, but I bet it outperforms Kestrel on VERY cheap and quick development, and cheap manufacturing.
-
That's what I figured too, but it's been running around my head and I realized battery+motor tends to do quite well on power to weight. Where it's not great is energy density for long duration. But a chemical rocket stage doesn't care about long duration, 3 minutes tops for the first stage. 3 minutes of 1 megawatt is 50 kilowatt hours, top of the line lithium-polymer might be able to do that in <200 kg. And the alternative likely isn't gas generator, it would be pressure fed for this size and simplicity of engine, which incurs a lot of structural penalty. And pressurized gas has even worse energy density. It's crazy but... the more I think about it, the cleverer I think this is.
Not sure whether these have enough power density, but since you don't need to recharge them, you can use a lithium primary battery. At 710 Wh/kg, your 50 kWh weighs just over 70 kg.
Incidentally there is an electric entry to the Pike's Peak hillclimb this year that has exactly these specs: 50 kWh and peak power 1020 kW.
http://www.tadiranbat.com/index.php/tadiran-batteries-better-by-design (http://www.tadiranbat.com/index.php/tadiran-batteries-better-by-design)
-
This seems like it would be good using solar on an upper stage or smaller batteries needed on an upper stage. Booster stage might take too much weight to power ratio.
-
This seems like it would be good using solar on an upper stage or smaller batteries needed on an upper stage. Booster stage might take too much weight to power ratio.
Would be interesting to do some trades. A big enough battery might win, maybe you could recharge off a small ICE a la ULA IVF. Because you're strictly limited in the amount of fuel you have I'm skeptical solar would win unless the endurance was quite long.
-
Another thought I had is, if this is an important improvement over pressure fed on small engines, it opens up a LOT of possibilities on eg lunar architectures. I bet electric pumped methalox would outperform pressure fed hypergols for example, without the extreme cost of RL-10, which is what you normally think of when you want moderate thrust and high ISP.
-
After looking further, the Tadiran cells have nowhere near the required power density.
For a booster stage we don't need to charge the battery ever, just needs to be cheap enough and light enough while providing the required power for the required time. And "cheap enough" here is just "cheaper than a turbopump" which isn't exactly a low bar.
-
I wonder if you could integrate a thermocouple into the chamber walls, serving as part of the power generation/cooling? Probably not enough to matter for the cost/complexity but who knows. I'm sure they have looked at it.
-
I wonder if you could integrate a thermocouple into the chamber walls, serving as part of the power generation/cooling? Probably not enough to matter for the cost/complexity but who knows. I'm sure they have looked at it.
Sounds like each engine is over a hundred kilowatts, it will be very difficult for anything other than the battery to contribute meaningfully to that.
-
- A generator going out on one engine of a cluster need not shut down that engine if other engines can spare the electricity.
What generator?
-
Thermoelectric generator materials do perform better with high deltaT, so if you could somehow line the inner layer of a regen nozzle's interior cooling passages, that miiiiiight work.
Some food for thought, is a pure electric turbopump necessarily the best? There has been recent automotive work on exhaust turbocompressors with colocated electric motor/generators on the same shaft. If the aim was more precise dynamic control of turbopump output, using a similar setup on a more conventional turbocompressor pump on a rocket might be interesting. Say, electric assist/control during normal ascent, but during maxQ throttledown, run the turbocompressor in generator mode to temporarily recharge the assist battery pack.
Left field idea is a twin counterrotating pack of carbon fiber flywheels for the power source, which could double as moment gyros by unbalancing the counterrotation. That can both absorb and output a stupid amount of power, can be left uncharged before flight, and can be charged/spun up just prior to flight using an umbilical (which you would probably already be using to dry spinup normal electric turbopumps at launch).
-
Riffing on that, if gas generators and associated turbine hardware doesn't scale down well, how about a single gas generator + electrical generator, then brushless DC driven pumps on each engine. Looks like Electron has a bunch of engines.
On the LI-ion front, just going over to hobbyking.com and playing with the LI-poly packs they have for sale. I believe we need a discharge rate of 20C based on the ratio of 1MW and 50kWH.
If you built a pack of 3400 of these:
http://www.hobbyking.com/hobbyking/store/__24602__ZIPpY
_Compact_1000mAh_4S_25C_Lipo_Pack_USA_Warehouse_.html
you end up with something like 50 kWH and a 1.25MW. At 105g each, that's 357kg, and includes the plug/wire/case on each battery. Cost is around $30k but I'm sure you could negotiate a discount :-)
Of course these batteries include margins enough that they can be used many times we only need a single good use. Chances are a specialized pack could cut the weight quite a bit.
-
37kw × 9 x 0.05 (3min 1/20th of hour) = 16.65kwhr for booster. NB that is assuming 37kw motor is running at 100% maybe a lot less.
For 2nd stage they could use multiple batteries and dump them as they are depleted.
-
Thermoelectric generator materials do perform better with high deltaT, so if you could somehow line the inner layer of a regen nozzle's interior cooling passages, that miiiiiight work.
Seems like the whole point of this design is extremely simple construction, high incremental complexity for low incremental performance seems like it's missing the point.
Some food for thought, is a pure electric turbopump necessarily the best? There has been recent automotive work on exhaust turbocompressors with colocated electric motor/generators on the same shaft. If the aim was more precise dynamic control of turbopump output, using a similar setup on a more conventional turbocompressor pump on a rocket might be interesting. Say, electric assist/control during normal ascent, but during maxQ throttledown, run the turbocompressor in generator mode to temporarily recharge the assist battery pack.
I don't think that's applicable here, no part of the turbomachinery is anywhere in any exhaust flow, so no way to generate any power from it. Start messing with that and I'm virtually certain you'll end up with a much more expensive and complex gas generator engine by the time you're done with the trades. Also, cars these days are chasing very small incremental efficiency gains.
-
Page for the engine:
http://www.rocketlabusa.com/about-us/propulsion/rutherford/
4600 lbf thrust
327s ISP
I think the ISP is deceptively good, as for a gas generator the propellant exhausted to power the pumps counts against the exhaust velocity, while for an electric pumped engine the power comes from batteries whose mass counts against the dry mass, and the same engine needs more battery the longer the burn. But, I think the best comparison is with pressure fed where similar considerations apply, and it looks like the first stage Rutherford has better ISP than the vacuum optimized Kestrel. Quite impressive.
Interview: http://www.popsci.com/rocket-labs-got-3d-printed-battery-powered-rocket-engine
"It’s really only the advancement in battery technology that has enabled us to go to electric turbopumps," Beck says. "Even three or four years ago, the technology wouldn’t have been sufficient."
Another example of electric technology defying expectations formed even quite recently.
-
327 ISP is that sea level or vacuum? Either way that about splits the isp of the RD-180 if its vacuum, if that's sea level then it exceeds it. Basically stage combustion level ISP without having to deal with stage combustion. Pretty impressive.
-
327 ISP is that sea level or vacuum? Either way that about splits the isp of the RD-180 if its vacuum, if that's sea level then it exceeds it. Basically stage combustion level ISP without having to deal with stage combustion. Pretty impressive.
I'm assuming it's vac because sl would require insane combustion chamber pressure to do that and that would incur a large battery mass penalty due to the increase in required pumping power. I'm pretty sure that refers to the first stage version since the given thrust times nine engines aligns with the thrust given for the overall vehicle.
If that's the first stage version, the vac version will probably beat its Merlin counterpart as well, so we might be looking at 350+ there. o_O
Basically these performance characteristics are what make me think of this as an alternative to pressure fed and expander, rather than large gas generators and staged combustion. It's simple and cheap like pressure fed, with the ISP of an expander, quite a large mass penalty but I suspect not the end of the world compared to pressure fed (since compressed gas has lower specific energy than battery) and MUCH simpler and more freedom in propellant choice than expander (is kerolox expander even possible?).
They're using it for both stages but this may have a lot of applications as a high energy US engine. Since it doesn't rely on the propellants for its power cycle and has a physically separate motor and pump for each, these should be easy to change, could use N2O+Propane, methalox, just about anything you can think of.
-
With the electric pump, you could do a trade and make it partially pressure fed. Since they are building the tanks out of carbon fiber, perhaps this allows them to use the combination of pressure and pump to get the required takeoff thrust, then throttle back later in flight so you don't run out of pressurant, battery power, or exceed g-load limits.
The "minimum gauge" issue might apply here -- CF tanks might require a minimum thickness, which in turn implies a certain tank pressure is possible. Add battery powered pumping on top of that to get to the chamber pressures you want.
-
As another data point, a PT6A-65 turboshaft has a dry weight of 220kg and shaft output around 1MW. But you'll need to add a generator to that, at which point the ~300kg of Li-poly batteries starts to look pretty competitive and a whole lot cheaper. Also that's going to consume some of your RP-1 and so impacts both dry weight and Isp.
-
With the electric pump, you could do a trade and make it partially pressure fed.
All LREs do this to an extent. Even just the hydrostatic pressure is probably >1atm on Electron for the LOX when fully loaded.
As another data point, a PT6A-65 turboshaft has a dry weight of 220kg and shaft output around 1MW. But you'll need to add a generator to that, at which point the ~300kg of Li-poly batteries starts to look pretty competitive and a whole lot cheaper. Also that's going to consume some of your RP-1 and so impacts both dry weight and Isp.
And more turbomachinery they have to develop.
-
What kind of pumps would you use for electric drive?
-
I wonder whether using these instead of pressure-feed in previously pressure-fed hypergolic systems would make a useful difference to performance (thinking improved mass fraction).
-
327 ISP is that sea level or vacuum? Either way that about splits the isp of the RD-180 if its vacuum, if that's sea level then it exceeds it. Basically stage combustion level ISP without having to deal with stage combustion. Pretty impressive.
...
Basically these performance characteristics are what make me think of this as an alternative to pressure fed and expander, rather than large gas generators and staged combustion. It's simple and cheap like pressure fed, with the ISP of an expander, quite a large mass penalty but I suspect not the end of the world compared to pressure fed (since compressed gas has lower specific energy than battery) and MUCH simpler and more freedom in propellant choice than expander (is kerolox expander even possible?).
....
I've sometimes wondered what is wrong the idea of having a 'tripropellant' kerolox rocket driven by an open-cycle hydrogen expander pump. (sorry off-topic)
-
Could this be used to pump fuel into the combustion-chamber of an oxidizer-rich staged combustion engine?
-
What kind of pumps would you use for electric drive?
Is the question as to types such as centrifugal, axial, and/or reciprocating pumps, for example? Motor characteristics may play a factor, such as high RPM, torque profile, matching to power source, AC/DC/polyphase, etc.
-
I've sometimes wondered what is wrong the idea of having a 'tripropellant' kerolox rocket driven by an open-cycle hydrogen expander pump. (sorry off-topic)
You mean like the RD-701 in kerolox mode, where kerosene is the primary fuel but hydrogen still fuels the turbopump? You can get some wicked chamber pressures and Isp with that. The RD-701 in kerolox mode had a 30 MPA chamber pressure and a 330s sea level Isp on a vaccum optimized engine. The vaccum isp was 415 s.
Somewhat off topic in this thread though.
-
Could this be used to pump fuel into the combustion-chamber of an oxidizer-rich staged combustion engine?
My intuition is that the high pressure gradient on SC would take too much power and enough battery that it wouldn't make sense. If you have staged combustion you have turbines powered by the preburner for pumping.
-
Ever since I heard of the concept I've always thought that electric pumps were a safer concept for amateurs building their first liquid rocket. Instead of having to build a heavy tank that can stand the pressures of blow-down and the risks of an exploding tank, you can use lightweight tanks and a commercially available pump running from mains power. I think amateurs often reject the idea because they don't think they'll be able to make a flight-weight pump-fed engine, and conclude they're going to need to master the skill of making high pressure talks anyway.
Even if your goal is to fly a pressure-fed vehicle, I think you can learn a lot about making rocket engines by starting an electric pump test stand. Of course, the big problem with this theory is that I've not found any such pump off-the-shelf which doesn't cost a fortune.
-
What kind of pumps would you use for electric drive?
Is the question as to types such as centrifugal, axial, and/or reciprocating pumps, for example? Motor characteristics may play a factor, such as high RPM, torque profile, matching to power source, AC/DC/polyphase, etc.
Good point.
John Whiteheads group at Livermore reckoned reciprocating positive displacement pumps (originally designed to replace pressure fed liquid systems) were better up to 5000 lb thrust.
Historically the bulk of US turbo pump systems have been centrifugal, with axial flow reserved for LH2.
Note there is no real reason for this beyond a)Von Brauns team built the V2 using pump designs derived from fire engine water pumps because they were worried about pressure pulses triggering combustion instability effects b)that's how Rocketdyne did it and as they were the US leaders in rocket engine production everyone played follow-the-leader.
Russian work went with radial inflow systems, which IIRC are actually a few percent more efficient, but a bit more difficult to design. OTOH they are somewhat easier to make once the design has been done.
In the US Barber Nicholls seem to be the leaders in this type of design and were (are ?) involved in the Merlin pump design for SX.
-
Does anyone remember a discovery called "Thermopower Wave"? The extremely high power demand of an electric turbopump sounds like a good fit for it:
http://www.technologyreview.com/news/530346/progress-on-a-powerful-new-way-to-generate-electricity/
http://www.pbs.org/wgbh/nova/next/tech/carbon-nanotube-fuses-unleash-surge-electrons-called-thermopower-wave/
http://www.nature.com/nmat/journal/v9/n5/abs/nmat2714.html
http://physicsworld.com/cws/article/multimedia/2013/feb/19/what-is-a-thermopower-wave
-
Any idea if it would be easier to 'play lego' with electro-pumps?
Kinda like the idea of a Falcon 5 1/3 with a penta-web of gas generator engines surrounding a small 1/3 thrust central electro-pump engine for improved throttle response when landing.
-
Does anyone remember a discovery called "Thermopower Wave"? The extremely high power demand of an electric turbopump sounds like a good fit for it:
No but if you read the first article they'd got to a 0.1% conversion efficiency, although that seems to be up from 0.000001% they started with.
This looks like something like thermionic generation (heat into direct free electron motion) and a "thermal battery" like the kind used by various weapon systems.
BTW Thermal batteries don't store heat. They are so called because they need heat to melt the (very) solid electrolyte of a primary cell and get the reaction started. Thermal batteries have "ignition sequences" to fire the incendiary mix that gets the battery working. This means they don't self discharge, giving very long shelf lives, at the expense of a) expense (built to the design of hardware they are powering with lead times to match) and b)a delay while they warm up.
I'd guess this tech will have similar issues, along with being about 1/1000 as efficient at present. :(
-
Out of curiosity; Anyone know how electromagneticly interactive cryogenically cooled fuels are? Like LOX and Liquid Hydrogen?
-
Since you don't need to recharge, a lithium sodium battery with twice the energy denisty at present levels of development seems applicable.
-
Could this be used to pump fuel into the combustion-chamber of an oxidizer-rich staged combustion engine?
Staged combustion is a way to burn some of the prop to run a turbopump.
Doesn't seem to be applicable to an electric pump engine.
Cheers, Martin
-
Could this be used to pump fuel into the combustion-chamber of an oxidizer-rich staged combustion engine?
Staged combustion is a way to burn some of the prop to run a turbopump.
Doesn't seem to be applicable to an electric pump engine.
Cheers, Martin
I meant using the turbopump to pump the oxygen (large volume) and the electric pump to pump the fuel (small volume assuming HC fuel) perhaps one small electric pump for the pre-burner and a bigger one for the main combustion chamber. thought it might simplify system design, while maintaining some of the power to weight of staged combustion.
-
Well, looking at a "typical" SC engine such as the shuttle, it has both low pressure and high pressure turbopumps. Maybe you could have a system where the low pressure pumps that are mostly there to pump the fuel could be electric, while the high pressure pumps that are for compression/cycle efficiency could be turbine driven?
-
Since you don't need to recharge, a lithium sodium battery with twice the energy denisty at present levels of development seems applicable.
exactly my thought, also Li (or Zn, Al)/Air (LOX?) come to mind
-
Single preburner turbine as electric generator, electrically coupled to oxidizer/fuel electric turbopumps, would allow shaft decoupling and temporary boost via secondary power sources. If you change your sizing for the preburner such that its generator output is matched to throttledown/maxQ inputs for the turbopumps, then a secondary battery can provide the boost assist at higher throttle settings with more fine tuning. I guess a hybrid electric staged combustion setup, where the preburner provides bulk electric power?
Err, so electric LP oxidizer/fuel turbopumps feed (electric HP oxidizer/fuel turbopumps + preburner power generator), plus the boost battery?