Author Topic: The Reaction Engines Skylon Master Thread (1)  (Read 566554 times)

Offline strangequark

  • Full Member
  • ****
  • Posts: 1074
  • Co-Founder, Tesseract Space
  • San Francisco, CA
  • Liked: 219
  • Likes Given: 12
Re: Skylon
« Reply #280 on: 06/20/2011 06:06 am »
baldusi asks about the propellant ratio.  For SABRE 3 in air breathing mode the equivalence ratio is 2.8 (at Mach 4). In rocket mode the mixture ratio is 6.

Am I mistaken in thinking that the stoichiometric ration is 8:1? I've seen how the RL10 has improved efficiency going from 5, to 5.5 in the A4 and 5.88 in the B2. The RS-25D, RD-0120 and RD-0146 have a ratio of 6 to 1, even the proposed Raptor had a ratio of 6:1. When I use the mole solution I get a ratio of 8 to 1. Yet, all current and new future high efficiency LH/LOX seem to converge to 6:1. What am I missing? In particular, wouldn't it improve the mass fraction in the skylon?

Stoich is 8. Isp is primarily a function of exhaust molecular weight and chamber temp. Chamber temp maxes out at stoich, but the molecular weight is also higher. Running it at O/F~6 decreases the molecular weight because you have excess light hydrogen, and it doesn't impact temp that much (the curve is pretty flat right around stoich), so it's a net gain.

Sorry to correct 93143 above, but the expansion ratio is actually independent of this consideration, and having stoich with an infinite expansion ratio would not be better than fuel rich.
« Last Edit: 06/20/2011 06:08 am by strangequark »

Offline Proponent

  • Senior Member
  • *****
  • Posts: 5815
  • Liked: 1309
  • Likes Given: 787
Re: Skylon
« Reply #281 on: 06/20/2011 06:24 am »
Stoich is 8. Isp is primarily a function of exhaust molecular weight and chamber temp. Chamber temp maxes out at stoich, but the molecular weight is also higher. Running it at O/F~6 decreases the molecular weight because you have excess light hydrogen, and it doesn't impact temp that much (the curve is pretty flat right around stoich), so it's a net gain.

Sorry to correct 93143 above, but the expansion ratio is actually independent of this consideration, and having stoich with an infinite expansion ratio would not be better than fuel rich.

I hadn't thought about this before, but I think 93143 is correct:  if the expansion ratio were infinite, stoichiometric would be the way to go.  Here's why.  The advantage of burning a hydrogen-rich mixture is not that the exhaust has lower molecular weight, it's that the exhaust contains a higher fraction of relatively simple molecules (hydrogen), which have fewer ways (principally rotational modes) of soaking up energy.  That means that for a given amount of expansion, more energy goes into bulk gas motion and less into internal modes.

At infinite expansion, however, the exhaust temperature falls to absolute zero and all energy is converted to bulk motion, regardless of the complexity of the molecules (in classical thermodynamics, of course; quantum reality would be a little different).  In that case, it's best simply to maximize the chemical energy per unit mass of the propellants, and that means a stoichiometric mixture ratio.
« Last Edit: 06/20/2011 06:28 am by Proponent »

Offline gospacex

  • Senior Member
  • *****
  • Posts: 3028
  • Liked: 540
  • Likes Given: 604
Re: Skylon
« Reply #282 on: 06/20/2011 06:28 am »
baldusi asks about the propellant ratio.  For SABRE 3 in air breathing mode the equivalence ratio is 2.8 (at Mach 4). In rocket mode the mixture ratio is 6.

Am I mistaken in thinking that the stoichiometric ration is 8:1? I've seen how the RL10 has improved efficiency going from 5, to 5.5 in the A4 and 5.88 in the B2. The RS-25D, RD-0120 and RD-0146 have a ratio of 6 to 1, even the proposed Raptor had a ratio of 6:1. When I use the mole solution I get a ratio of 8 to 1. Yet, all current and new future high efficiency LH/LOX seem to converge to 6:1. What am I missing?

8:1 runs hotter and exhaust has more unreacted oxygen, which is chemically aggressive. IOW: harder on engine chamber and throat.

Offline Proponent

  • Senior Member
  • *****
  • Posts: 5815
  • Liked: 1309
  • Likes Given: 787
Re: Skylon
« Reply #283 on: 06/20/2011 06:43 am »
Its not just engine efficiency that counts.   The tanking mass for the vehicle increases a lot when using larger ratio's of LH2 to LOX, due to LH2's very low density.


I might have understood it backwards, but the mixture is usually oxidized to fuel. So a mixture of 8:1 would be better than a 6:1. LH2 is about 16 lighter than LOX. So 8:1 would mean 15% less tank volume.

Look at it this way: a rocket engine converts chemical energy into thrust in two steps.  First, it burns propellants in the combustion chamber to generate heat.  Then in the nozzle it converts that heat into kinetic energy in the form of exhaust gases shooting out the back.  The efficiency of the engine is the product of the efficiencies of the two stages.  The efficiency (per unit mass of propellant) of combustion is maximal at the stoichiometric ratio of 8:1.  The efficiency of the nozzle is maximal with a ratio of 0:1 (pure hydrogen, a ratio at which the efficiency of combustion is alas zero).  Maximal efficiency of the engine is a compromise between the two; I'll take 93143's word for it that it's 4 or 5 to one.  Maximal efficiency of the vehicle, as kraisee and 93143 mention, will be pushed toward a higher ratio (less hydrogen) by hydrogen's low density.
« Last Edit: 06/20/2011 06:47 am by Proponent »

Offline 93143

  • Senior Member
  • *****
  • Posts: 3038
  • Liked: 294
  • Likes Given: 1
Re: Skylon
« Reply #284 on: 06/20/2011 09:05 am »
I'll take 93143's word for it that it's 4 or 5 to one.

I should be more careful when I state stuff like that.

...go ahead and take it, for the sake of argument, but bear in mind that it's based on what I remember of some CEA results I generated a while back.  CEA will provide theoretical rocket performance for either frozen composition or equilibrium composition, and you have to interpolate to find the true performance.  The interpolation weighting will of course change with expansion ratio and mixture ratio, but the only method I can think of to pick a weighting is to try to match the performance of an existing engine, so I can't easily track that variance...

Sadly, IIRC the frozen and equilibrium curves do not peak at the same value, or anywhere near it, and the interpolated curve is pretty flat in between.

In other words, if someone has access to better data, I'm willing to be corrected.  But I very much doubt that the maximum Isp will occur above 6:1 for any halfway plausible expansion ratio...

Online Lampyridae

  • Veteran
  • Full Member
  • ****
  • Posts: 1878
  • Liked: 227
  • Likes Given: 450
Re: Skylon
« Reply #285 on: 06/20/2011 01:07 pm »
Stoich is 8. Isp is primarily a function of exhaust molecular weight and chamber temp. Chamber temp maxes out at stoich, but the molecular weight is also higher. Running it at O/F~6 decreases the molecular weight because you have excess light hydrogen, and it doesn't impact temp that much (the curve is pretty flat right around stoich), so it's a net gain.

Sorry to correct 93143 above, but the expansion ratio is actually independent of this consideration, and having stoich with an infinite expansion ratio would not be better than fuel rich.

I hadn't thought about this before, but I think 93143 is correct:  if the expansion ratio were infinite, stoichiometric would be the way to go.  Here's why.  The advantage of burning a hydrogen-rich mixture is not that the exhaust has lower molecular weight, it's that the exhaust contains a higher fraction of relatively simple molecules (hydrogen), which have fewer ways (principally rotational modes) of soaking up energy.  That means that for a given amount of expansion, more energy goes into bulk gas motion and less into internal modes.

At infinite expansion, however, the exhaust temperature falls to absolute zero and all energy is converted to bulk motion, regardless of the complexity of the molecules (in classical thermodynamics, of course; quantum reality would be a little different).  In that case, it's best simply to maximize the chemical energy per unit mass of the propellants, and that means a stoichiometric mixture ratio.

I'd add that you also run less chance of bathing the engine in superhot oxygen if something goes wrong. The Russians do have some lox-rich engines though IIRC, fancy alloys and all that.
SKYLON... The League of Extraordinary Gentlemen's preferred surface-to-orbit conveyance.

Offline Proponent

  • Senior Member
  • *****
  • Posts: 5815
  • Liked: 1309
  • Likes Given: 787
Re: Skylon
« Reply #286 on: 06/20/2011 01:34 pm »
I'm sure it's entirely impractical, but is there any possibility that the performance of a hydrolox rocket stage would be optimized by burning a stoichiometric mixture of hydrogen and oxygen diluted with liquid helium -- or, since this is a purely theoretical discussion anyway -- liquid helium-3?  The helium atom is heavier than the hydrogen molecule, but I would expect it to have essentially only translation degrees of freedom in this application.
« Last Edit: 06/20/2011 01:50 pm by Proponent »

Offline baldusi

  • Senior Member
  • *****
  • Posts: 7444
  • Buenos Aires, Argentina
  • Liked: 1461
  • Likes Given: 4554
Re: Skylon
« Reply #287 on: 06/20/2011 06:03 pm »
Well, may be the solution would be to run the chamber at 8:1 and use TAN with pure H2?

Offline strangequark

  • Full Member
  • ****
  • Posts: 1074
  • Co-Founder, Tesseract Space
  • San Francisco, CA
  • Liked: 219
  • Likes Given: 12
Re: Skylon
« Reply #288 on: 06/21/2011 01:17 am »

I hadn't thought about this before, but I think 93143 is correct:  if the expansion ratio were infinite, stoichiometric would be the way to go.  Here's why.  The advantage of burning a hydrogen-rich mixture is not that the exhaust has lower molecular weight, it's that the exhaust contains a higher fraction of relatively simple molecules (hydrogen), which have fewer ways (principally rotational modes) of soaking up energy.  That means that for a given amount of expansion, more energy goes into bulk gas motion and less into internal modes.

At infinite expansion, however, the exhaust temperature falls to absolute zero and all energy is converted to bulk motion, regardless of the complexity of the molecules (in classical thermodynamics, of course; quantum reality would be a little different).  In that case, it's best simply to maximize the chemical energy per unit mass of the propellants, and that means a stoichiometric mixture ratio.

Is that actually true that you would recover the vibratory energy in bulk motion, even moving down to 0K? I thought there were internal damping mechanisms for molecular vibration. In that case, and given the infinite amount of time to reach infinite expansion, there should be losses that scale as a function of molecular complexity.

It is an interesting question though, and perhaps I was wrong to so quickly dismiss it for the infinite expansion case. Though, what I said should be true for any finite nozzle of arbitrary size, correct?

I'd add that you also run less chance of bathing the engine in superhot oxygen if something goes wrong. The Russians do have some lox-rich engines though IIRC, fancy alloys and all that.

Yes and no. They run ox-rich (fuel-lean) in the preburners, to produce "low"-temperature gases to run the turbines (low temperature meaning maybe 1800F). The preburner gas is extremely ox-rich, maybe an O/F of 30, but is therefore much cooler than the combustion chamber.

There are no kerolox engines of which I'm aware that run fuel-lean in the main combustion chamber. Stoich for kerolox is O/F~3.4, and the Russian engines are typically about 2.7 or so for the MCC. If you were to run an engine at say 3.8, then you'd have 7000F gas with about 15% free oxygen, which is a real good way to fry your engine, fancy alloys or not.

« Last Edit: 06/21/2011 01:31 am by strangequark »

Offline ciscosdad

  • Full Member
  • *
  • Posts: 169
  • Liked: 16
  • Likes Given: 178
Re: Skylon
« Reply #289 on: 06/21/2011 01:27 am »
Helium 3 is an interesting idea. I think the best part is the understanding we will get from its effect on the processes. I've always been fascinated by the actual process that converts the chemical energy of the propellent to the kinetic energy of the vehicle, and the influence of O/F ratio is at the heart of that.


Cost may be a tiny issue though :-\

Offline Proponent

  • Senior Member
  • *****
  • Posts: 5815
  • Liked: 1309
  • Likes Given: 787
Re: Skylon
« Reply #290 on: 06/21/2011 04:38 am »
Is that actually true that you would recover the vibratory energy in bulk motion, even moving down to 0K? I thought there were internal damping mechanisms for molecular vibration. In that case, and given the infinite amount of time to reach infinite expansion, there should be losses that scale as a function of molecular complexity.

I think you're right that there are a number of idealizations needed in making the infinite-expansion nozzle 100% efficient.  As some point as the gas expands, the timescale for equilibration among the various modes becomes so long that effectively the energy can no longer be extracted from, for example, rotation.

Quote
Though, what I said should be true for any finite nozzle of arbitrary size, correct?

I think the result is that as the nozzle efficiency increases (i.e., as the expansion ratio increases toward the point at which the exit pressure equals the ambient pressure), combustion efficiency becomes relatively more and more important.  Hence, the optimal mixture ratio will shift closer and closer to stoichiometric as the expansion ratio increases.
« Last Edit: 06/21/2011 04:45 am by Proponent »

Offline Proponent

  • Senior Member
  • *****
  • Posts: 5815
  • Liked: 1309
  • Likes Given: 787
Re: Skylon
« Reply #291 on: 06/21/2011 04:45 am »
Well, may be the solution would be to run the chamber at 8:1 and use TAN with pure H2?

I don't know a lot about TAN, but off hand it seems to me that you'd want to mix the additional working fluid (excess hydrogen) in as early as possible, because 1) it can then increase the efficiency of the expansion process at all stages rather than only after some of the expansion has occurred, and 2) it will lower the chamber temperature and reduce the risk of any engine-threatening oxidizer-rich spots appearing.

On second thought, I can think of one advantage of mixing the excess hydrogen in later (TAN-style):  since the hydrogen would be injected at lower pressure, less pump power would be required (the power needed by a pump is proportion to the pressure differential times the volume flow rate).  That would leave more power for the engine itself, resulting in higher specific impulse.  I'd think, though, that this would be a pretty small advantage.

Offline aceshigh

  • Full Member
  • ****
  • Posts: 665
  • Liked: 197
  • Likes Given: 16
Re: Skylon
« Reply #292 on: 06/21/2011 07:53 pm »
Hemsell... this question is not exactly is not exactly about Skylon, but about the somewhat related A2 Lapcat.

many people doubt hipersonic (or even supersonic) aircraft will exist in the next decades because of the economics of fuel... Boeing and Airbus will spend billions to make airplanes that are more economical rather than being faster or anything.

how much $$ of fuel would an A2 Lapcat spend per km compared with a common 747?

Offline Hermit

  • Member
  • Posts: 42
  • Liked: 1
  • Likes Given: 1
Re: Skylon
« Reply #293 on: 06/22/2011 06:48 am »
Can any further discussion not relating to SKYLON or Reaction Engines work be taken to another thread.

Offline Hempsell

  • Full Member
  • *
  • Posts: 145
  • Liked: 62
  • Likes Given: 0
Re: Skylon
« Reply #294 on: 06/22/2011 01:53 pm »
Have you considered advanced carbon carbon for the aeroshell? It flew on the x37 and is meant to be only a quarter the density of RCC. It is also stronger and more oxidant resistance than RCC.

No we haven’t thanks for the steer – we will be relooking at TPS towards the end of the year.

Offline Hempsell

  • Full Member
  • *
  • Posts: 145
  • Liked: 62
  • Likes Given: 0
Re: Skylon
« Reply #295 on: 06/22/2011 01:55 pm »
I would rather like that, but as the passenger module is not part of the SKYLON development plan it won’t be up to us.

Interesting. So REL wouldn't design a baseline passenger module, but would leave that up to the customers (unless, presumably, they hire you do it)? Sounds like a reasonable conservative approach.

How involved will REL be in the design of the ground facilities?

This is absolutely the case, not only for passenger modules but also upper stages and the other support equipment mentioned in the User Manual Annexes, They are guides as to the sort of thing SKYLON can support.  We find without these concept studies people did not realize SKYLON can carry people and launch geostationary satellites and build stations.

One thing that might change is that the ESA report did suggest the upper stage should be included in the SKYLON development.  We understand why they made this recommendation and it is being considered.

Another point to made regarding this question and by later question by tnphysics is that these support systems do not have to look like our concept designs.  It is up to the poeple developing them to design something that meets their requirements, be they commercial or governmental.

With Spaceports as far as possible we will supply the interfaces requirements and it is up to the spaceport and the SKYLON operators to provide the facilities.  In reality it will not be this clean cut and some of the SKYLON support facilities will have to be provided by the SKYLON constructor but we are looking for independent (and competing) spaceports and they may want to offer facilities to competing launch system to SKYLON, so we do not want to make them tied franchises.

Offline Hempsell

  • Full Member
  • *
  • Posts: 145
  • Liked: 62
  • Likes Given: 0
Re: Skylon
« Reply #296 on: 06/22/2011 01:56 pm »
Re passenger module: Annex D of the Skylon User's Manual describes a conceptual Personnel/Logistics Module that would accommodate 24 people.

http://www.reactionengines.co.uk/downloads/SKYLON_User_Manual_rev1-1.pdf

It sounds as though the current 'D' Skylon designs will have more internal space than when this was written, so perhaps more capacity for passengers.

No - the D design will have the same size and capabilities as outlined in the User Manual.  The point of the User Manual was to publicly ask the question are these the right capabilities for SKYLON?  Given nobody has said they were not the right capabilities, we are now using this as the starting point for the D1 design.  There maybe some minor improvements to the detailed interfaces as we progress but the size and mass in the User Manual is what you are going to get.

Offline Hempsell

  • Full Member
  • *
  • Posts: 145
  • Liked: 62
  • Likes Given: 0
Re: Skylon
« Reply #297 on: 06/22/2011 01:57 pm »
Hemsell... this question is not exactly is not exactly about Skylon, but about the somewhat related A2 Lapcat.

many people doubt hipersonic (or even supersonic) aircraft will exist in the next decades because of the economics of fuel... Boeing and Airbus will spend billions to make airplanes that are more economical rather than being faster or anything.

how much $$ of fuel would an A2 Lapcat spend per km compared with a common 747?

I think it is too early to say – remember LAPCAT is only a technology advancement project looking into the long term.  Our design was one of a number of concepts used to find areas where technology development could answer questions on feasibility and cost.  However the provisional look at our A2 design suggested that if the seat cost were comparable to current business class fares it would be economic. But again I emphasise these are very early assessments.

Offline aceshigh

  • Full Member
  • ****
  • Posts: 665
  • Liked: 197
  • Likes Given: 16
Re: Skylon
« Reply #298 on: 06/22/2011 02:49 pm »
thanks Hempsell.

@Hermit: sorry pal. Wont do it again. But anyway, A2 is kinda related to Skylon... not only it looks similar (but larger and white :)) as
"The Scimitar engines use related technology to the company's earlier SABRE engine, which is intended for space launch, but here adapted for very long distance, very high speed travel."

Offline Matt32

  • Member
  • Posts: 15
  • Liked: 0
  • Likes Given: 0
Re: Skylon
« Reply #299 on: 06/23/2011 11:14 am »
Some more skylon snippets here:

http://www.rocketeers.co.uk/node/1595

Apparently there are plans for ten X-37 sized nacelle test vehicles to be constructed...
« Last Edit: 06/23/2011 11:47 am by Matt32 »

Tags: