Author Topic: Electric pumps in rockets, MHD generators  (Read 5966 times)

Offline savuporo

  • Senior Member
  • *****
  • Posts: 5137
  • Liked: 946
  • Likes Given: 338
Electric pumps in rockets, MHD generators
« on: 09/10/2016 08:50 PM »
Jon Goff mentioned MHD generators in the RocketLab thread, which i thought is a novel thing worth discussion.

Obviously there is some prior work, like this paper looking at various MHD generator configurations in scramjet applications.
Comparison of Generator Performance of Small-Scale MHD Generators with Different Electrode Dispositions and Load Connection Systems , 2014, Toru Takahashi, Takayasu Fujino, and Motoo Ishikawa, University of Tsukuba. They have a set of related papers.

Digging around more, NTRS has this paper
Design study - Rocket based MHD Generator, 1997, ERC Inforporated

Yet another paper on similar subject
Analysis of MHD Generators for use with solid rocket motors, 2016, Technical University of Munich

Is this a thing that could be realistically made to work, to power electric pumps for liquid biprop rockets ? Any better and more solid references ?

EDIT: Found the Russian SAKHALIN reference.
http://mhd.ing.unibo.it/Old_Proceedings/1999_Beijing/Beijing%201999/VOLUME2/Vol2Cap01.pdf
"Pulsed MHD Power System SAKHALIN-The World Largest Solid Propellant Fueled MHD Generator of 500MWe Electric Power Output"

« Last Edit: 09/10/2016 08:54 PM by savuporo »
Orion - the first and only manned not-too-deep-space craft

Offline Bob Shaw

  • Full Member
  • ****
  • Posts: 956
  • Liked: 355
  • Likes Given: 331
Re: Electric pumps in rockets, MHD generators
« Reply #1 on: 09/10/2016 09:30 PM »
The USSR was big on MHD in the past; if there were real prospects of rocketry applications then I'd expect that their current-day counterparts would be trying to push this technology.
« Last Edit: 09/10/2016 09:56 PM by Bob Shaw »

Offline savuporo

  • Senior Member
  • *****
  • Posts: 5137
  • Liked: 946
  • Likes Given: 338
Re: Electric pumps in rockets, MHD generators
« Reply #2 on: 09/10/2016 09:51 PM »
The USSR was big on MHD in the past..

Apparently, yes. I did quick reading up on history of MHD generator research, and looks like at some point there were serious bets on achieving 60-70% efficiencies in utility scale power generation. Sample down here, from this paper. Research has been done in US, Italy, Japan, Russia.. India and China as well. There used to be even IEEE MHD power generation conferences.

EDIT: and to amend, most of the current generator research seems to be focused on hypersonic flight applications, again in multiple separate programs across the world.
« Last Edit: 09/10/2016 09:54 PM by savuporo »
Orion - the first and only manned not-too-deep-space craft

Offline Bob Shaw

  • Full Member
  • ****
  • Posts: 956
  • Liked: 355
  • Likes Given: 331
Re: Electric pumps in rockets, MHD generators
« Reply #3 on: 09/10/2016 10:00 PM »
In the mid 1960s there were serious proposals to run combat RADAR stations off of solid-fuel rocket motors running MHD generators - this was covered a couple of times in the BIS Spaceflight magazine at the time (1966 rings a bell). I think this morphed into torpedo applications.

Offline savuporo

  • Senior Member
  • *****
  • Posts: 5137
  • Liked: 946
  • Likes Given: 338
Re: Electric pumps in rockets, MHD generators
« Reply #4 on: 09/11/2016 12:07 AM »
And because this is kind of specific to electric pumps,
 
A high performance, electric pump-fed LOX / RP propulsion system, Phase II Project, 2014-2016, Ventions LLC
http://techport.nasa.gov/externalFactSheetExport?objectId=18314
Orion - the first and only manned not-too-deep-space craft

Offline lamontagne

  • Full Member
  • ****
  • Posts: 954
  • Liked: 1273
  • Likes Given: 215
Re: Electric pumps in rockets, MHD generators
« Reply #5 on: 09/11/2016 06:46 AM »
MHD was looked at as a topping application for coal power plants, and needs to run very hot.  MHD first, then turbine.  The efficiency is combined efficiency, not MHD alone.

The combined gas cycle for natural gas, using a gas turbine feeding a steam turbine achieved similar efficiencies, without the complications (lots of erosion at those temperatures, need to have a conducting gas).  So MHD was pretty much abandoned.

« Last Edit: 09/11/2016 06:47 AM by lamontagne »

Offline Hobbes-22

  • Full Member
  • ***
  • Posts: 349
  • Acme Engineering
    • Acme Engineering shop
  • Liked: 98
  • Likes Given: 74
Re: Electric pumps in rockets, MHD generators
« Reply #6 on: 09/11/2016 11:36 AM »
MHD introduces extra weight and extra conversion steps into the chain to drive a turbopump.

Compare:
preburner exhaust->turbine->shaft->turbopump impellers
preburner exhaust->MHD generator->electric motor->shaft->turbopump impellers

Per Wikipedia, MHD efficiency is on the order of 30%. A turbopump can achieve 70-80% efficiency.

For a large engine, you need lots of power. The Vulcain 1 needed 3.7 MW for the LOX turbopump and 11.9 MW for the LH pump. Electric motors that powerful aren't small.

Offline Burninate

  • Full Member
  • ****
  • Posts: 1129
  • Liked: 343
  • Likes Given: 72
Re: Electric pumps in rockets, MHD generators
« Reply #7 on: 09/11/2016 01:09 PM »
MHD introduces extra weight and extra conversion steps into the chain to drive a turbopump.

Compare:
preburner exhaust->turbine->shaft->turbopump impellers
preburner exhaust->MHD generator->electric motor->shaft->turbopump impellers

Per Wikipedia, MHD efficiency is on the order of 30%. A turbopump can achieve 70-80% efficiency.

For a large engine, you need lots of power. The Vulcain 1 needed 3.7 MW for the LOX turbopump and 11.9 MW for the LH pump. Electric motors that powerful aren't small.

Off on a tangent, but -
* Superconducting electric motors are much smaller/lighter, provided they recieve coolant.
* Deep cryocoolant in hydroLOX engines is free
* Moderate temperature cryocoolant in methane-LOX or RP1-Lox is free for HTS materials.

Offline savuporo

  • Senior Member
  • *****
  • Posts: 5137
  • Liked: 946
  • Likes Given: 338
Re: Electric pumps in rockets, MHD generators
« Reply #8 on: 09/11/2016 05:56 PM »
I dont think anyone will dispute that you will get better performance ( thrust to weight, ISP ) out of a turbopumped engine. And it's very likely that scaling up electric pumping will make little sense.

That being said, electrically pumped engines might very well find their niche, for things like Electron and maybe for deep space workhorse engines where reduced plumbing complexity may be a benefit. ( Note that there has not yet been a pumped biprop that has been fired in space after long duration, only pressure-fed )

So the question is, can MHD generation somehow optimize or improve a fully battery powered electric pump ? I think its clear that you'd still have to have a battery, at least for starts and restarts. But you might be able to downsize it from full-duration power delivery to maybe 1/4th or 1/8th of that, given that battery specific power specs can deliver this.

So this brings the question, if this can be made to work for example on a LOX/hydrocarbon biprop. Where do the electrodes go ? What and where gets injected, metals ?

Off on a tangent, but -
* Superconducting electric motors are much smaller/lighter, provided they recieve coolant.
So most of the development for high specific power motors seems to be going on in electric aviation. NASA and others have been doing lots of research, and then there are actual real world accomplishments.

Siemens flew a 260kw, 50kg air-cooled motor in "Extra 330LE" aerobatic plane. So 5KW/Kg. NASA research from a few years ago peaked with 10KW/Kg with motors submerged in LN2. I don't think i've seen higher projections than that.

EDIT: Nah, HTS-motors are around 20.0 kW/kg apparently now, or higher. Source: Brown, G. V, “Weights and Efficiencies of Electric Components of a Turboelectric Aircraft Propulsion System,” 2011
« Last Edit: 09/11/2016 06:51 PM by savuporo »
Orion - the first and only manned not-too-deep-space craft

Offline Shevek23

  • Member
  • Posts: 37
  • Liked: 1
  • Likes Given: 0
Re: Electric pumps in rockets, MHD generators
« Reply #9 on: 09/14/2016 07:23 PM »
MHD introduces extra weight and extra conversion steps into the chain to drive a turbopump.

Compare:
preburner exhaust->turbine->shaft->turbopump impellers
preburner exhaust->MHD generator->electric motor->shaft->turbopump impellers

Per Wikipedia, MHD efficiency is on the order of 30%. A turbopump can achieve 70-80% efficiency.

For a large engine, you need lots of power. The Vulcain 1 needed 3.7 MW for the LOX turbopump and 11.9 MW for the LH pump. Electric motors that powerful aren't small.

Reading over savpuro's post, I think you have this backwards. The idea would not be to have any kind of pre burner or gas generator; it would be to take MHD power off the exhaust of the main combustion chamber. The MHD generator is on top of the nozzle. As I understand the original post, it would in effect make the rocket engine more similar to a turbojet, where the turbine extracts the mechanical power needed to drive the compressors.

To modify your text diagram:

turbopump impellers->main combustion chamber->MHD generator->nozzle->Thrust {propellant flow}
                 ^                                                              ||
                 ||                                                              V
electric motor <=               electric current                         {electric power flow--upstream!}

I hope that shows up more or less aligned in people's displays! The idea is to omit separate pre burners and mechanical turbines and feed directly off the main thrust flow. This will cool the exhaust and thus reduce thrust but all propellants flow through the chamber and are exhausted in the main drive stream, and all combustion takes place in the main chamber alone.

Conceptually it is like an expansion cycle engine, except power to drive the pumps does not come from auxiliary cooling loops but from the main thrust generation process. Expansion cycles are hard to scale up due to cube/square law issues, but this cycle scales with the power available in the main combustion chamber outlet flow, which is to say proportional to thrust so it could scale up indefinitely.

Since pumping power is only a small fraction of the total power released by the main combustion process, it is quite all right if MHD has a theoretical efficiency limit of 30 percent or so; we want to use a lot less than that, as little as possible, just what we need to run the electric powered pumps and any other auxiliary power we need.

One reason an expansion cycle is elegant is that we use power flow from a loop we require in order to keep the nozzle from melting anyway. Regenerative cooling recycles nozzle/chamber waste heat one way or another, by the heated propellant (usually fuel) feeding in to pre burners, gas generators, or the expansion turbine. Going with an MHD generator instead means the coolant loop has to feed directly into the main chamber, but the heat is still recycled so we don't lose anything there. We have eliminated auxiliary combustion chambers of any kind.

Now your point about the weight of electric motors is, well, weighty, and presumably applies to the MHD generator too.

Against this, first of all are electric motors really much heavier than the alternatives we can actually use? Other methods all require high temperature gas turbines of some kind. J-2S used chamber tap-off to supply the turbine for the pumps, but had to dilute it with extra fuel (hydrogen) to cool it down to ranges that turbines could handle. I would guess that advances in high-temperature turbine engines over the past half century would allow a leaner and meaner version today but we still would take a hit in diverting fuel flow, and another hit in that we are tapping off the main chamber thus weakening it. Gas generators divert a portion of the fuel & oxidant to generate the power-the turbine in that system may well be far lighter than an equivalent power and RPM electric motor, but have you factored in the need for the gas generator, associated plumbing, and the diversion of loaded propellant mass around the main thrust cycle? Could it be that even if gears are included, modern electric motors can be competitively light? Especially if we use superconducting magnets which as someone pointed out above, we can keep cool with the flow of hydrogen on the way to the hot elements, or for high temp superconductors we can use the oxygen flow?

Of course the motor to drive the pumps is only half what the OP proposed to add: we need the MHD generator too. But we don't need it to extract more than a few percent of the available power in the exhaust!

All the Wikipedia stuff is talking about using MHD as one stage in a staged electrical power generation process, and generally is talking about using coal for fuel too.  But IIRC early MHD research did use rocket engines. The gas temperatures and velocities of typical rocket exhaust (at chamber outlet) would be much higher than a coal dust combustion in air output would be, so on one had conventional mechanical turbines would be vaporized, but the conditions for MHD to work well are more naturally attained. Think of it as a virtual turbine then, and acknowledging that it is heavier than what a theoretical mechanical turbine would be if only we had materials to make one from, consider we are drawing off the tremendous main power flow, not some auxiliary side flow. I'd think that rocket exhaust, at the chamber mouth anyway if not at nozzle exit, is already a plasma, or partially so, enough to allow extraction of the power flow we'd need to drive the pumps. If not--seeding it with extra material to give it the needed charge might be bad if we inject it along with unconsumed propellant and it interacts with the chamber reaction and perhaps does bad things materially to the chamber, by chemical corrosion or formation of soot or the like. But what if we can inject it just upstream of the generator? Will the charge-creating additives have time to do their work? If they can, they are probably mostly harmless to the thrust generation process in the nozzle--they add a tiny bit of mass after all.

For appropriate sized engines, expansion cycle seems to me to provide most of the advantages this electrical cycle might. But those engines have to use a gas type fuel like hydrogen or methane, and as noted they are size limited. I'd like to see the numbers on state of the art electric motors and realistic MHD setups to generate the electricity--plus of course the conductive wiring needed to efficiently bring this power back up to the pumps. Note that if we need superconducting materials to bring the weights down, these are an option with cryogenic propellants involved. Then there is the matter of price to consider. Assuming all rocket engines are throw-away, use once only, it may be that with or without superconductors, electric pumping is too expensive considering we discard all the kit involved. But again that is a matter of real world numbers, not any theoretical considerations. In some ways the MHD-driven-motor concept can be far simpler and more reliable, and also controllable, than the ingenious all-mechanical systems we use. They might well be superior if we choose to develop ways to economically reuse engines.

I'm not so sure that even if they do tend to be substantially heavier than mechanical components (even bearing in mind all the auxiliary stuff the realistic mechanical versions require, such as gas generators or pre burners) that that is such a drawback. The engine is only a small part of a typical rocket, even of just the dry mass of that rocket. Having a lower thrust/weight ratio in the engine might be acceptable if offset by other advantages.                                                             

Offline Jim

  • Night Gator
  • Senior Member
  • *****
  • Posts: 31222
  • Cape Canaveral Spaceport
  • Liked: 9494
  • Likes Given: 298
Re: Electric pumps in rockets, MHD generators
« Reply #10 on: 09/14/2016 07:40 PM »
In some ways the MHD-driven-motor concept can be far simpler and more reliable, and also controllable, than the ingenious all-mechanical systems we use.
                                                       

There is no way that is true, see the H-1.
a.  how do you start a MHD-driven-motor?  All it takes a turbine engine is head pressure, simple start cartridge, or gas bottle. 
b.  The number of parts in an H-1 are minimal for RP-1 engine.  Only way to get simpler is to be a pressure fed engine.
c.  The H-1 could be refired after a dunk in salt water.

Gas generators are among the lightest parts of an engine.
« Last Edit: 09/14/2016 07:42 PM by Jim »

Offline Jim

  • Night Gator
  • Senior Member
  • *****
  • Posts: 31222
  • Cape Canaveral Spaceport
  • Liked: 9494
  • Likes Given: 298
Re: Electric pumps in rockets, MHD generators
« Reply #11 on: 09/14/2016 07:44 PM »

Against this, first of all are electric motors really much heavier than the alternatives we can actually use? Other methods all require high temperature gas turbines of some kind. J-2S used chamber tap-off to supply the turbine for the pumps, but had to dilute it with extra fuel (hydrogen) to cool it down to ranges that turbines could handle. I would guess that advances in high-temperature turbine engines over the past half century would allow a leaner and meaner version today but we still would take a hit in diverting fuel flow, and another hit in that we are tapping off the main chamber thus weakening it. Gas generators divert a portion of the fuel & oxidant to generate the power-the turbine in that system may well be far lighter than an equivalent power and RPM electric motor, but have you factored in the need for the gas generator, associated plumbing, and the diversion of loaded propellant mass around the main thrust cycle? Could it be that even if gears are included, modern electric motors can be competitively light?                                                           

Yes, electric motor and pumps would be heavier than any gas generator engine.  Rocket turbo pumps are the highest power per mass devices.

And with MHD, you are ignoring that it will take energy out of the gas flow.  It will be a parasitic power draw much like a gas generator.
« Last Edit: 09/14/2016 07:47 PM by Jim »

Offline pippin

  • Regular
  • Senior Member
  • *****
  • Posts: 2523
  • Liked: 276
  • Likes Given: 39
Re: Electric pumps in rockets, MHD generators
« Reply #12 on: 09/14/2016 07:54 PM »
Gas generators may be very simple but turbines are not and they are especially pretty hard to manufacture given the tolerances (if you want them lightweight).
Electrical pumps do have their merits in simplicity even if weight might not be their prime advantage.

Offline Jim

  • Night Gator
  • Senior Member
  • *****
  • Posts: 31222
  • Cape Canaveral Spaceport
  • Liked: 9494
  • Likes Given: 298
Re: Electric pumps in rockets, MHD generators
« Reply #13 on: 09/14/2016 07:57 PM »
 
Regenerative cooling recycles nozzle/chamber waste heat one way or another, by the heated propellant (usually fuel) feeding in to pre burners, gas generators, or the expansion turbine.
                                                         

wrong, the heat picked up the propellant is not wasted.  Propellant that going into the nozzle goes into the combustion chamber.  Propellant for pre burners and gas generators is tapped off before going into the nozzle.

 Propellant going to an expansion turbine after the nozzle is using the heat energy to turn the turbine.  It is not wasted and in fact, this means all the propellant will be burned in the main chamber.  That is why this type engines have some of the highest ISP.
« Last Edit: 09/14/2016 07:59 PM by Jim »

Offline Jim

  • Night Gator
  • Senior Member
  • *****
  • Posts: 31222
  • Cape Canaveral Spaceport
  • Liked: 9494
  • Likes Given: 298
Re: Electric pumps in rockets, MHD generators
« Reply #14 on: 09/14/2016 07:59 PM »
Gas generators may be very simple but turbines are not and they are especially pretty hard to manufacture given the tolerances (if you want them lightweight).
Electrical pumps do have their merits in simplicity even if weight might not be their prime advantage.

for small engines I agree

Offline savuporo

  • Senior Member
  • *****
  • Posts: 5137
  • Liked: 946
  • Likes Given: 338
Re: Electric pumps in rockets, MHD generators
« Reply #15 on: 09/14/2016 09:43 PM »
There is no way that is true, see the H-1.
a.  how do you start a MHD-driven-motor?  All it takes a turbine engine is head pressure, simple start cartridge, or gas bottle. 
You'd still need a battery, a high specific power one, and as small as possible to transition to steady state. Fortunately, decent high power batteries with respectable specific energy do exist for a while now.

If this would be a deep-space engine attached to a spacecraft, specifically designed for very long loiter times, you'd have a battery in the spacecraft anyway.

Again, nobody has yet fired a turbopump in space, after any significant loiter times.


Gas generators may be very simple but turbines are not and they are especially pretty hard to manufacture given the tolerances (if you want them lightweight).
Tolerances, plus the thermal environment in turbine. Cryogenic temps on one side, hot burning gases on the other.

Yes, electric motor and pumps would be heavier than any gas generator engine.  Rocket turbo pumps are the highest power per mass devices.
Most definitely - especially if you consider engine alone. However, as an integrated spacecraft system, where battery weight can be shared, it might not be so bad.
« Last Edit: 09/14/2016 09:48 PM by savuporo »
Orion - the first and only manned not-too-deep-space craft

Offline Jim

  • Night Gator
  • Senior Member
  • *****
  • Posts: 31222
  • Cape Canaveral Spaceport
  • Liked: 9494
  • Likes Given: 298
Re: Electric pumps in rockets, MHD generators
« Reply #16 on: 09/14/2016 10:01 PM »

Again, nobody has yet fired a turbopump in space, after any significant loiter times.


Not an issue for the RL10

Offline Jim

  • Night Gator
  • Senior Member
  • *****
  • Posts: 31222
  • Cape Canaveral Spaceport
  • Liked: 9494
  • Likes Given: 298
Re: Electric pumps in rockets, MHD generators
« Reply #17 on: 09/14/2016 10:07 PM »

If this would be a deep-space engine attached to a spacecraft, specifically designed for very long loiter times, you'd have a battery in the spacecraft anyway.


There are better solutions for those.   Such a spacecraft is not going to have non storable propellants.


Offline savuporo

  • Senior Member
  • *****
  • Posts: 5137
  • Liked: 946
  • Likes Given: 338
Re: Electric pumps in rockets, MHD generators
« Reply #18 on: 09/14/2016 10:35 PM »
Again, nobody has yet fired a turbopump in space, after any significant loiter times.
Not an issue for the RL10
Maybe for some future version or mission kit. But so far, IIRC the record longest loiter time stands at 11h or so. Not going to the Moon and back with that.
Orion - the first and only manned not-too-deep-space craft

Offline Shevek23

  • Member
  • Posts: 37
  • Liked: 1
  • Likes Given: 0
Re: Electric pumps in rockets, MHD generators
« Reply #19 on: 09/14/2016 10:44 PM »
Regenerative cooling recycles nozzle/chamber waste heat one way or another, by the heated propellant (usually fuel) feeding in to pre burners, gas generators, or the expansion turbine.
                                                         

wrong, the heat picked up the propellant is not wasted.  Propellant that going into the nozzle goes into the combustion chamber.  Propellant for pre burners and gas generators is tapped off before going into the nozzle.

 Propellant going to an expansion turbine after the nozzle is using the heat energy to turn the turbine.  It is not wasted and in fact, this means all the propellant will be burned in the main chamber.  That is why this type engines have some of the highest ISP.

Read what I said--I said regeneration uses the waste heat. Either it goes into the combustion chamber directly, warmed thus recycling the heat that must be removed from the nozzle lest it melt, or it goes into an expansion turbine--which cools it in extracting useful pump energy. With a pre burner to give the turbo pump more oomph, there is some burning of the fuel to make it hotter before the turbine, but we are still keeping the heat extracted below. Thus as I said "recycles nozzle/waste heat."

Well and good. If MHD current generation were feasible, which is not an absolute and for all time, but a function of current state of the art, of course it is going to reduce the thrust. If we take 10 percent of of the kinetic energy out of a moving stream it winds up moving at 95 percent of the speed it was before so of course that's a 5 percent reduction in thrust and ISP.

With an expansion cycle this does not happen because the energy driving the turbo pump would have been wasted otherwise, being heat leaking from the reaction and flow of generated thrust gas. But a loss does show up in that in extracting power from the mere expansion of some fuel under this heat, the exhaust from the turbine is both cooled and lowered in pressure. The latter means we have to pump this flow of coolant/expansion fluid/fuel to a higher pressure to drive it through the plumbing and then have enough after we take some power out of it to enter the main combustion chamber, so we pay 2 ways--higher pumping load and lower injection temperature. All the mass of propellant does go through the combustion chamber. But we lose some thrust we theoretically could have had if we could pump the propellant for free.

I do not scorn the H-1, or other gas generator rockets. (I hear good things about it on another site, and have even incorporated it into a little project of my own). But they too, obviously, penalize the thrust versus a magic system that is pumped for free. In this case because some propellant is diverted from the main chamber flow completely and vented overboard with little propulsive effect--it is instead purposed to generate the necessary power. As I understand it the advantage of this kind of open cycle gas generator is that the auxiliary generator chamber pressure (and temperature) can be modest because one is exhausting into relatively low pressure--at worst, one atmosphere, dropping to effectively zero as the rocket ascends. So the amount of fuel thus diverted is modest. Nevertheless, in figuring the ISP of the rocket as a whole, one must add the propellant mass diverted to the mass divisor, though it produces no thrust directly to add to the dividend thrust, and thus we have lower ISP and less thrust than the same amount of propellant might have got us if we could only get massless elves to crank the fuel pump for us.

Same thing for staged combustion; now all the propellant mass does go to the chamber, but it is somewhat depleted, diluted with partial exhaust products. And the plumbing is greatly complicated too, and all has to be at pressure exceeding what is needed to inject it into the combustion chamber to account for losses in all the steps involved, and I believe the temperature at the turbine inlets has to be considerably higher than open-cycle gas generators.

Expansion cycle is elegant and simple but limited--either one pumps the fuel (always hydrogen, which is hard to pump, although it might work with or other volatile hydrocarbons too) to extra high pressure so it has enough to enter the main chamber after being worked by the turbine, or else one ejects the cooling fuel flow to space after a low-pressure turbine extracts more of its energy. Limited by the square-cube law in that there is only so much heat to pick up in the nozzle, and the bigger the engine, the area goes up by the square of linear dimension but for a given chamber pressure, the volume of fluids to pump goes up as the cube. Hence such engines are generally pretty low thrust.

Basic physics says if you need to pump up propellants to a given chamber pressure, it costs power to do so and that power, barring magic, comes from somewhere in the system. So it is absurd to say tapping it off with an MHD generator imposes a penalty without observing that every other possible system must also do so.

When Wikipedia gives a figure like 30 percent for the efficiency of MHD, they mean that if you have a gigawatt of power in the exhaust from a chamber/nozzle system, MHD can't extract more than 300 MW from it. It does not mean that for every watt you extract, you somehow render useless another 2 watts in the exhaust; I don't see how the basic physics of using electromagnetic interactions to generate electromotive force and hence current would do that at all. If you need an order of magnitude less than 300 MW, just 30 or so for your pumps, you only need to take that much, leaving 97 percent of the energy still there and available for 98.5 percent of the thrust you'd get from the same chamber and nozzle without MHD impeding it.

It may well be that a mechanical turbine driving pumps with a shaft will lose less useful power than MHD emf driving current through wires to an electrical motor, but the electric system will still deliver most of the power usefully, whereas it is quite impossible, as I've seen you yourself point out with a minimally informative "NO," to tap the power of the exhaust from the main chamber--because, as you did not bother to elaborate, the dang turbine would be vaporized. Therefore all the efficient turbines you cite are elsewhere, running on weaker flows that require extra plumbing and pumping the MHD system does not require. If we assume 10 percent of the power the MHD generator taps from the main exhaust is lost as resistance or hysteresis heating, well that is a lot of heat to be sure--but in that case the same fuel flow that cools the nozzle can be somewhat augmented, with modest pressure and volume flow increases, to first cool the motor, transmission lines and generator before flowing on into the nozzle, from which, greatly heated and mostly by the nozzle, it goes into the main chamber. Since we need to pump coolant through the nozzle anyway, the only thing you can charge against this alternate system as far as extra pumping goes is the path through the electrical stuff, which will add very little. Versus either a completely separate if low pressure/temperature gas generator (vulnerable also to fluctuations in power due to varying exterior pressure) or high pressure and temperature staged turbines.

All of that costs. You are saying it costs little, and that must be so or rockets wouldn't work at all, but why then should we doubt electric motors and MHD generators might also cost relatively little in mass or cost as well?

I have no reason to think, in the current state of the art, that existing MHD generators or even electric motors are competitive with the kinds of turbo machinery that astronautics has developed up to date. Of course not; no one has been investing any effort in it. Not directly. But the original post points out some people who are working on improving MHD generation for other purposes, and meanwhile there are people working on lighter, more powerful electric motors--Tesla Motors for instance.

Meanwhile of course the state of the art of turbo machinery has also advanced, largely for aeronautical purposes, but one supposes that if someone had to replicate the basic plumbing of the H-1, or F-1A, gas generators today but were free to incorporate advances in materials engineering since the early 60s they'd tighten up the generator, making it run hotter with less diversion of fuel and oxidant through a much hotter running turbine, driving lighter and more effective pumps. To get the lightest, hottest version possible they'd have to spend a lot of money of course, and I daresay one reason the H-1 engine took being dunked in seawater was that wasn't so ultra-hot, so if we wanted to make it reusable instead of lighter and more efficient (by virtue of diverting less propellant to the gas generator system) we'd use only modestly improved materials, or the same ones, with no superheating it and mainly saving money while having an even more robust machine.

So it may well be that MHD generation/electric motor driven pumps are a dead end, simply because of facts about which approach is lighter and cheaper. Perhaps down to the fundamental physics of electric motors, which certainly do tend to work on principles requiring substantial masses of magnetic material, or perhaps just because turbo machinery has pulled ahead in the development race.

But what I saw in this thread was people gainsaying a concept they did not seem to properly understand the basic layout of because they did not read it very carefully, disingenuous comparisons that charged the new concept with costs they overlooked the old ones have to pay as well, and cheap shot gotchas.

Offline jongoff

  • Recovering Rocket Plumber/Space Entrepreneur
  • Senior Member
  • *****
  • Posts: 6019
  • Lafayette/Broomfield, CO
  • Liked: 1993
  • Likes Given: 681
Re: Electric pumps in rockets, MHD generators
« Reply #20 on: 09/15/2016 02:24 AM »
In some ways the MHD-driven-motor concept can be far simpler and more reliable, and also controllable, than the ingenious all-mechanical systems we use.
                                                       

There is no way that is true, see the H-1.
a.  how do you start a MHD-driven-motor?  All it takes a turbine engine is head pressure, simple start cartridge, or gas bottle.

You start such an engine using power from batteries that are sized to run the engine long enough for the MHD power to start running the pumps and refilling the batteries. Since they only need to run for maybe 3-10s, and possibly at a lower throttle setting, they don't need to be that big. Could also be ultracapacitors.

~Jon
« Last Edit: 09/15/2016 02:24 AM by jongoff »

Online Robotbeat

  • Senior Member
  • *****
  • Posts: 26862
  • Minnesota
  • Liked: 6768
  • Likes Given: 4800
Re: Electric pumps in rockets, MHD generators
« Reply #21 on: 09/15/2016 02:42 AM »
In some ways the MHD-driven-motor concept can be far simpler and more reliable, and also controllable, than the ingenious all-mechanical systems we use.
                                                       

There is no way that is true, see the H-1.
a.  how do you start a MHD-driven-motor?  All it takes a turbine engine is head pressure, simple start cartridge, or gas bottle.

You start such an engine using power from batteries that are sized to run the engine long enough for the MHD power to start running the pumps and refilling the batteries. Since they only need to run for maybe 3-10s, and possibly at a lower throttle setting, they don't need to be that big. Could also be ultracapacitors.

~Jon
...for the first stage, could be started with ground power.

Here I'll propose something a little crazy, too:

Hot start the second stage using power generated by the first.

(But batteries work pretty well.)
Chris  Whoever loves correction loves knowledge, but he who hates reproof is stupid.

To the maximum extent practicable, the Federal Government shall plan missions to accommodate the space transportation services capabilities of United States commercial providers. US law http://goo.gl/YZYNt0

Offline Shevek23

  • Member
  • Posts: 37
  • Liked: 1
  • Likes Given: 0
Re: Electric pumps in rockets, MHD generators
« Reply #22 on: 09/15/2016 02:57 AM »
I dont think anyone will dispute that you will get better performance ( thrust to weight, ISP ) out of a turbopumped engine. And it's very likely that scaling up electric pumping will make little sense.....
It may or may not always be a loser versus turbine power overall in terms of thrust to weight; the low weight of established turbine systems, despite their added (but low-power, relative to the main engine thrust* speed) bric-a-brac probably is hard to beat.

However, ISP should not be a problem. That is a function of fuel mix, pressure, and chamber/nozzle design and ambient conditions, none of which relate directly to powering the pumps. Or rather, traditional methods do relate to it; gas generators burn additional propellant on the side which should be deducted from the net engine ISP, and so on as said above. Tapping the power off the main nozzle output should only cause an ISP hit comparable to what other methods cost, one way or another.

It occurs to me that if it is necessary to seed the flow with some ionizing substance, that mass flow counts against ISP too, so if it is large compared to the basic mass flow, say 10 percent, it does cost some ISP. In so doing it would also raise the thrust of course by adding more molecules to the mix--cooling down everything else, but its own momentum counts against the reduction of the other stuff. ISP goes down but thrust goes up. Then it all takes the hit due to withdrawal of power by the magnetic field retarding the flow a bit.

But how much?

Look at the H-1. It delivered 947 kN in vacuum, at ISP of 289 seconds. Let's upgrade the thrust up to an even 1000, one million Newtons or a bit over 100 metric tons equivalent force. The power of the stream is thus 1x10^6 x 289*9.81=2.835 gigaWatts! The mass flow is presumably the thrust divided by the effective velocity, or 352.72 kg/sec, of which 109.2 is kerosene type fuel and the other 243.52 is oxygen. These would have volume of 128.5 and 211.75 liters respectively, or 0.34 cubic meters total. Each of these has to be pressurized to 40 atmospheres, from say 1.5 in lightly pressurized storage; 1 ATM is about 1x10^5 Pascals and so the power requirement is about 1310 kW. Say that with one thing and another we really need 2 megaWatts, overall then we need just 1/1418 of the total power the main chamber produces. If it were an expansion driven engine and the heat loss to the nozzle were 1 percent of that total (the truth is the exhaust has even more energy per second than nearly 3 GW because it has a lot of uncaptured energy in the form of heat and turbulence, and it is the heat that drives the loss that heats the nozzle, so I can only guess just how great that flow is) and we can thermodynamically only use half of it to drive a turbine--well, there is still 7 times the power we need available so our turbine could lightly cool and depressurize the coolant flow and still, if it were modestly over pressurized say to 50 ATM coming out of the pump, be injected into the chamber to burn. But of course an H engine is not expansion driven, it runs off a gas generator. Now if it were possible for that gas generator to use the same mix and pressures as the main chamber, then the mass of diverted propellant we'd need would be tiny--each kg bears over 8 megaJoules of recoverable energy, so just a quarter kg/sec would be more than enough. If I assume this has been included in figuring the engine ISP of 289, then actually the ISP of what comes out of the nozzle would be 289.2, and the energy in the nozzle flow would be that much greater--same thrust times slightly higher velocity means about 2 more megawatts--in fact, the pumping power is added to the total. But it only gets us 289 ISP overall because of that extra quarter-kg a second we have to dispose of.

2 megawatts is a lot of power on a human scale to be sure. But if we are extracting it from a 2.837 GW flow, it only slows the exhaust down by one meter/sec, or reduces the ISP from ideal 289.2 to 289.1. Thus we'd come out ahead of the gas generator approach on that score! By a minuscule amount of course.

Now you say:
EDIT: Nah, HTS-motors are around 20.0 kW/kg apparently now, or higher. Source: Brown, G. V, “Weights and Efficiencies of Electric Components of a Turboelectric Aircraft Propulsion System,” 2011

In that case, we ought to be able to drive the pumps with a 100 kg motor. To be sure, we must also generate the power in the first place; at this point I have no idea how to estimate what a 2 MW MHD generator drawing power off a 2.4 GW flow ought to mass; another 100 kg seems like a fair guess at the minimum. Being pessimistic, let's say it needs to be 300 kg all up, and that before adding in the mass of the actual pumps the motor drives. Would it be reasonable to guess that a gas generator and turbine that delivers 2 MW shaft power to those same pumps masses 50 kg? If so we are worse off by 250 kg per engine! Encyclopedia Astronautica gives the dry weight of the H-1 as 635 kg so with another 250 it would be 885, a 40 percent worsening that reduces the thrust weight ratio from over 150 to a mere 109. With 8 on a Saturn 1B we have added 2 tons to the stage dry mass.

But we also add 2766 Newtons of thrust in vacuum; this is only 1/7 what we lose in dead weight. But I am only guessing at total masses. It could be better, might be worse. And despite the deserved praise of the tremendous power/weight ratio of turbine systems, perhaps the gas generator did not mass a mere 50 kg? If it massed 100, then the only added weight is the MHD generator and wiring. Since the MHD generator works off very rapid fluid flow I can well believe it might be a lot lighter than the motor. We keep the tiny thrust/ISP gain regardless so we might break even on T/W or even come out a tiny bit ahead.

I note you have not suggested installing megawatt scale motors for booster engines like the H series. You suggested instead that they might address challenges to scaling motors down, for deep space applications. There  the RL-10 is the one to beat or improve, if we want a high ISP hydrogen/oxygen mix. It is more difficult to pump low-density and volatile liquid hydrogen, but as Jim points out the RL-10 manages it pretty well. The pumping power would be considerably greater per unit mass flow--but of course that is true for the turbine driven systems also.

Jim says it is a poser how one would ever start an electrically pumped engine. It does seem a bit problematic for high-powered systems that restart--but the H-1, his shining example, could not practically be restarted in flight either. Starting an electric version on the ground has the very obvious solution of running electric cables to each engine (or to one heavy current bus on the stage connecting all the pump motors) and give it current from outside the rocket. Once the motors spin up and the chambers are ignited the engines obviously run on MHD power.

Suppose we did want to restart in flight? Batteries to store that kind of energy--more to the point, deliver it at such high power rates--would indeed be problematic to say the least--though if we could devise some kind of capacitor that could hold the charge for a very long time, we could always restart because we could recharge it during a burn. And if we have a cluster of engines--like the Saturn 1B's 8 for instance--we could shut down some of them and restart them later by drawing current off the running engines.

Agena developed a system of bellows to enable restart of its hypergolic engines as many times as desired; it should be possible for a deep-space engine like RL-10 to include reserve tanks of hydrogen and oxygen that we keep both in at very high pressures, to spin up the pumps (supplemented by a trickle of modest battery power) and flood the chamber, say to partial pressure, and ignite that, which would restore power flow from the MHD generators.

Jim dismissed the problem of throttling control, but insofar as it does after all remain tricky for turbine powered pumps, we could have some advantage from electric pumping, since electricity is very easy to control. As with turbine driven systems there is feedback between the actual chamber/nozzle flow--not the case with gas generator systems where the drive for the turbine is independent of the chamber flow to be sure, but there too the rate of control of the pumping speed is governed by the speed-of-sound ruled physics of the gas generator, as is also the case with staged combustion (further complicated there because chamber conditions do affect the turbines, as the chamber is their output). Here, bearing in mind that we only want to capture a tiny fraction of the power of the exhaust stream, we ought to be well able to either increase that power draw by amplifying the magnetic field a bit and thus overcome any tendency of the engine to fade, or vice versa pull back sharply on the pump in case some feedback situation has it running away, at speeds more like the speed of light than the (very fast, at many thousands of degrees to be sure) speed of sound. A good control system ought to be able to keep the engine running very smoothly since we simply apply the prescribed amount of power needed microsecond to microsecond. We ought to be well able to prevent issues like pogo, or anyway counter them very effectively.

Which suggests to me that we might want to have hybrid systems, wherein we use one of the other turbine powered strategies, but have a fairly strong motor mounted on the turbine axis to assert direct control over its gyrations, and power it with a light MHD generator; the electrical system would exist to overpower surges and transients the mechanical system might be subject to. As such it can be considerably lighter than one that needs to do the whole job. Considering that if we can rely on the electrical system for say 10-20 percent of sustained power, we can lighten the turbo-system accordingly, and thus perhaps offset any added mass the electrical system brings, or anyway keep the growth of the whole system within bounds.

Finally, it occurs to me also there might be more than one way to get the power out of the main chamber work. Instead of attempting to draw power from the entire exhaust stream, we could have tap-offs of the main chamber, mini-nozzles with very closely concentrated but small in dimension MHD systems that attempt to capture not 1/1000 of total power from all the exhaust, but say 1/10 each from 4 tap-offs, collectively drawing off just 1 percent of the flow. Note that if we take 10 percent from each, each one still has 95 percent the ISP and thrust. So we have 99 percent plus .95 of 1 percent of the thrust. This is a tiny bit less thrust than we'd get by tapping all the power off the whole flow, but it still comes out ahead in ISP terms of the gas generator solution. I suggest using many taps rather than one mainly for reasons of thrust balance; it also provides some redundancy should something go wrong with one of the tap MHD generators, since maximum MHD efficiency--that is, capacity to extract energy from a given flow--is 30 percent, but we are designing around 10 but with a huge surge capacity; one tap out means the remaining three need to draw 13.3333 percent instead; even with two out 20 percent ought to be within capacity for such small streams. If it turns out to be necessary to expand the flow to full velocity to get the power efficiently, or to salt it heavily with some sort of charge-enhancing substance, doing so with just 1 percent of the flow instead of all of it seems very advantageous.

Offline Shevek23

  • Member
  • Posts: 37
  • Liked: 1
  • Likes Given: 0
Re: Electric pumps in rockets, MHD generators
« Reply #23 on: 09/15/2016 03:12 AM »
...
(But batteries work pretty well.)

Well, we all know that batteries are pretty poor in specific energy per mass, also they suffer breakdown and fatigue especially if one attempts to draw massive currents from them all at once. If a capacitor could hold the charge for hours, days, even weeks, then that would be the way to go, but I believe any such thing would drain over time.

There are ways to use the gas-charge methods beloved of turbine engine designers though. I had the notion of having tap-off MHD generators running on just 1 percent or so of the main chamber flow; it might be possible to design an alternate feed for these mini-nozzles and run them briefly off hypergolic or solid charges, or even take the main propellants, with small quantities stored under somewhat above main chamber pressure. One might devise half-domes that extend over the dimple within the main chamber where each tap-off is, creating a temporary mini-chamber, and feed a trickle of the main gases plus any seeding substances needed to generate the power needed to spin up the pumps and pressurize the main chamber. Then, withdraw these "eyelid" covers with the min-chamber stuff still hot to ignite the main chamber. We'd probably still need some battery power to do this, hopefully we can charge a capacitor gradually and then discharge it quickly to prevent surge draws on the battery. But it would be a much lighter battery than we'd need to strain to fill the whole chamber and ignite it.

Flywheels might also serve, if not for long-term storage, than medium term--batteries slowly spin up the wheels, powerful regenerative braking of the wheels quickly charges the capacitor, and a surge of pump power is available as needed to smooth over any gaps or roughness in the start.

And pressure starting does not only work on turbines. The Hindenburg's diesel engines were not started electrically but with surges of compressed air to the pistons; in the same way it should be possible to introduce high-pressure jets of oxygen and fuel into the respective turbo-pumps so as to spin them up (assisted by a relatively modest stored electric current to the motors, which can regulate the spin up as well) and a pre-pressurized supply can flow into the engine chamber while this is happening.

The chamber, especially with the engine in vacuum, will not need to start at full pressure either. Once it starts the MHD generators come into play and can bootstrap the main pumps all the way up.

This might be a bad thing to try on the ground but on the ground, power from off the rocket is readily available.

Offline Proponent

  • Senior Member
  • *****
  • Posts: 5104
  • Liked: 768
  • Likes Given: 526
Re: Electric pumps in rockets, MHD generators
« Reply #24 on: 09/19/2016 06:38 AM »

Digging around more, NTRS has this paper
Design study - Rocket based MHD Generator, 1997, ERC Inforporated

Attached is the wonderfully concise 1971 Russian paper (in English) that underlies the ERC report.

Also attached is a straight-forward 1960 paper from the Rand Corporation that provides a simple introduction to rocket-powered MHD.

Offline savuporo

  • Senior Member
  • *****
  • Posts: 5137
  • Liked: 946
  • Likes Given: 338
Re: Electric pumps in rockets, MHD generators
« Reply #25 on: 09/01/2017 02:35 AM »
Siemens flew a 260kw, 50kg air-cooled motor in "Extra 330LE" aerobatic plane. So 5KW/Kg. NASA research from a few years ago peaked with 10KW/Kg with motors submerged in LN2. I don't think i've seen higher projections than that.

Came across an interesting related article and seems appropriate to post here.

http://cafe.foundation/blog/electric-propulsion-for-big-birds/

Quote
Ambitious Goals, Different Approaches
Goals are ambitious, with NASA Research Agreements (NRAs) awarded to the University of Illinois and Ohio State University to develop electric systems that can achieve 13 kilowatts per kilogram and efficiency greater than 93 percent.  NASA Glenn’s target is 16 kW/kg and 98-percent efficiency.

General Electric and the University of Illinois share an NRA to make power converters that produce 19 kW/kg and an efficiency target of 99 percent.  Boeing’s working on a cryogenic converter with goals of 26 kW/kg and an efficiency of 99.3 percent.  Compare these goals with the Energy Department’s 2020 goal of 14.1 kW/kg for vehicle power electronics.
Orion - the first and only manned not-too-deep-space craft

Offline john smith 19

  • Senior Member
  • *****
  • Posts: 5861
  • Everyplaceelse
  • Liked: 741
  • Likes Given: 4489
Re: Electric pumps in rockets, MHD generators
« Reply #26 on: 09/01/2017 07:53 AM »

Again, nobody has yet fired a turbopump in space, after any significant loiter times.


Not an issue for the RL10
Not proven.

I don't think an RL10 mission has gone over 24 hours, and most a good deal below 12.

AFAIK anything above that has been pressure fed hypergols
"Solids are a branch of fireworks, not rocketry. :-) :-) ", Henry Spencer 1/28/11  Averse to bold? You must be in marketing."It's all in the sequencing" K. Mattingly.  STS-Keeping most of the stakeholders happy most of the time.

Offline john smith 19

  • Senior Member
  • *****
  • Posts: 5861
  • Everyplaceelse
  • Liked: 741
  • Likes Given: 4489
Re: Electric pumps in rockets, MHD generators
« Reply #27 on: 09/01/2017 08:02 AM »
So Electron has shown battery driven pumps are viable for small LV's but this is a different thing.

The pumps are possible. The Motors may have an issue with heat, but if one of the propellants is a cryogen you have a heat sink available.

I dimly recalled this needed a magnetic field but it seems you can do MHD without fields which leaves the question how conductive is the exhaust, and how heavy the conductors to collect the power?

Right now this is TRL 0 tech.

As for sticking this on the back end of a SCramjet you've got to be kidding.
"Solids are a branch of fireworks, not rocketry. :-) :-) ", Henry Spencer 1/28/11  Averse to bold? You must be in marketing."It's all in the sequencing" K. Mattingly.  STS-Keeping most of the stakeholders happy most of the time.

Offline Asteroza

  • Full Member
  • ****
  • Posts: 468
  • Liked: 51
  • Likes Given: 2
Re: Electric pumps in rockets, MHD generators
« Reply #28 on: 09/01/2017 08:48 AM »
Ultimately you have to look back at what is providing the shaft power to the pump. Assuming an electric motor, then where is the electrical power of the right type coming from? Changing MHD electrical output to motor input is not going to be easy.

The HVEPS program did do some practical experiments/demonstrations in supersonic MHD generation (ostensibly for future USAF scramjet platforms...) but the mag fields are nontrivial. Size issues would mean you would need a magnet right after the nozzle throat (or the suggestion of tapping off main chamber flow to MHD specific side channels/nozzles), and probably a mechanically advanced wire/stick in the injector head to push seed material into the chamber for MHD.

There was RIME, which was using a generator rocket for MHD generation to feed MHD acceleration of a flow, so similar conceptually...

Rocket-Induced Magnetohydrodynamic Ejector 

Offline Asteroza

  • Full Member
  • ****
  • Posts: 468
  • Liked: 51
  • Likes Given: 2
Re: Electric pumps in rockets, MHD generators
« Reply #29 on: 10/03/2017 02:26 AM »
There was a recent post elsewhere that pointed out 2014 work on a seedless MHD generator design in a scramjet inlet context, the takeaway being they think e-beam based plasma non-equilibrium ionization is viable. (Not UV laser based?)

https://www.netl.doe.gov/File%20Library/Events/2014/MHD/2-1-MPGW-NASA-IBlanksonPresentation10012014.pdf

The segmented hall generator design they show though would be tough to make as a combustion chamber/throat/nozzle unless the chamber is effectively segmented to support the vanes, but perhaps may be amenable to a dual bell/aerospike shaped nozzle where there is a center post-like artifact.


Tags: