Author Topic: The Reaction Engines Skylon Master Thread (5)  (Read 419631 times)

Offline lkm

  • Full Member
  • ****
  • Posts: 508
  • Liked: 91
  • Likes Given: 1
Re: The Reaction Engines Skylon Master Thread (5)
« Reply #540 on: 08/15/2015 04:56 PM »
The use of altitude compensating Expansion/Deflection nozzles should help also.

It's only an expansion deflection nozzle in airbreathing mode, in rocket mode it's a 120:1 expansion ratio bell nozzle. Either way an E-D nozzle will still become under expanded once it hits it's critical pressure and the plume will grow.

Hmmm, I think your saying the E-D benefit occurs it lower altitudes and so is if no help here; that makes sense.

But the nozzle doesn't know if SABRE is in air-breathing or rocket mode, right? Maybe you simply meant earlier in flight when you said 'It's only an expansion deflection nozzle in airbreathing'?

The exhaust from the bypass burners (air-breathing mode) will affect the plume, but that's turned off by the time we're at the problematic speed & altitude/pressure. 

Not quite. The SABRE 4 cycle works by separating the rocket combustion chamber from the airbreathing combustion chamber allowing a high pressure rocket chamber (170 bar) and a low pressure airbreathing chamber ( 12 bar but can be from 6 bar to 20 bar depending on implementation). Each rocket  nozzle consists of a single rocket chamber surrounded  by 3 ( or more depending on implementation) airbreathing chambers, they share a rocket nozzle by having a 30:1 ratio inner nozzle after the rocket chamber throat and then a closable throat fed by the airbreathing chambers followed by a nozzle extension that takes the rocket expansion ratio to 120:1. The upshot being that in airbreathing mode the nozzle acts an E-D nozzle but in rocket mode the airbreathing throat closes and the rocket sees a conventional bell nozzle ( although they are designed to work concurrently to allow for a smooth transition).
My suggestion was to  move to a single large area ratio nozzle and use the airbreathing throat to do TAN as altitude compensation. My second suggestion is to crank the delta wing to increase the wingspan, my third is to consider alternate fuels as the SABRE 4 should allow the rocket mode to work on different fuels to the airbreathing mode and gelled hydrogen could be used for the airbreathing mode, together fuel volume could be slashed creating  a smaller tail.
« Last Edit: 08/15/2015 04:56 PM by lkm »

Offline john smith 19

  • Senior Member
  • *****
  • Posts: 5859
  • Everyplaceelse
  • Liked: 741
  • Likes Given: 4489
Re: The Reaction Engines Skylon Master Thread (5)
« Reply #541 on: 08/15/2015 05:05 PM »
I wouldn't be so sure. All the heat flow is in the boundary shock wave and their indications of more than 10 times the free stream temperature... However the final gas model will be (they do give compensation estimates for real gas), this is massive, we are talking about a kelvin-scale here!
You're missing a few points.

Firstly the vast bulk of that exhaust will be in the plume. A very small part of it's mass flow will be hitting the aft fuselage.

Secondly this is a point simulation and Skylon is a on launch trajectory. It's outside air pressure is continually falling. That means the plume is growing but also the plume density is falling, lowering the energy of the gas that will hit the aft fuselage.

Being a point simulation you'd miss the integration of temperature involved. You've also missed that a large part of the body is at a much lower temperature and therefor that heat could be moved by a heat pipe arrangement. Such systems were looked at for the Shuttle wing but never flown.

Like the XS 1 you have to ask how long does the fuselage have to sustain that temperature. On a M5 cruise vehicle that's hours, but the whole trajectory for Skylon is 10-20min tops.
Quote
They will probably have to cool the whole body aft of the wing, I doubt you can passively cool this without adding a lot of mass.
The report is not detailed enough to make that assumption.
Quote
This essentially means there's a significant chance the whole airframe concept is not feasible. It's the kind of risks you run into when designing all-new stuff and venturing into unknown unknowns with a lot of elements in your design.
Depends on your assessment of how much unknown you're dealing with.

We already know the Shuttle can survive reentry from these kinds of temperatures. No their TPS would not be the preferred option, but an existence proof for survival in horizontal landing from orbital velocity is quite encouraging.

OTOH we didn't know you couldn't build even a reasonably large 2nd stage VTOL stage to survive reentry until someone actually tried it and found they couldn't make it work.

BTW doing a CFD for this size of model takes a lot of computer time, even today. Other posters can probably put some numbers on how long a 6-8 million point model takes to run.

It's a tribute to how far Skylon's design has been matured by REL that they had a CAD model to supply to NASA for this work. I doubt NASA would have looked doing this study a decade ago, given this is a non US vehicle that is not already flying.

Moving to real gas chemistry multiplies that amount of time, so operating at this level of detail is good enough for a (relatively) quick look at the problems.

I will note a few points about the work in general.

1) The Nacelle is fully  closed at anything above M5.5.
2) Likewise ramjet flow (and any burner induced drag) ends when the nacelle closes
3) They use a SSME nozzle, but IIRC that's 1:77, not 1:100
4) It's not an E/D, although above M5.5 I'm not sure there would be much difference.
5) Nozzle form is important. Hempsell alluded to this in a Space Show appearance. A NATO report pointed out that Russian nozzles are differently designed and less prone to flow separation. I'm not sure if the British use the Russian or the Western approach.
6)It's not a "turbo ramjet," that's basically the J58 on an SR71, and a very different thermal
cycle.  IIRC Bill Escher's taxonomy described it as a "Deeply pre-cooled air turbo rocket."

Work like this establishes boundaries for further work, narrowing the risk space. It also shows how subtle the problems can be with a design and puts some boundaries on how serious they are.
"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 lkm

  • Full Member
  • ****
  • Posts: 508
  • Liked: 91
  • Likes Given: 1
Re: The Reaction Engines Skylon Master Thread (5)
« Reply #542 on: 08/15/2015 05:24 PM »
Just for reference, the System 2 material can take a maximum of 1470K and the expected maximum temp it should see was 1100K.

Offline adrianwyard

  • Full Member
  • ****
  • Posts: 947
  • Liked: 160
  • Likes Given: 209
Re: The Reaction Engines Skylon Master Thread (5)
« Reply #543 on: 08/15/2015 05:25 PM »
The use of altitude compensating Expansion/Deflection nozzles should help also.

It's only an expansion deflection nozzle in airbreathing mode, in rocket mode it's a 120:1 expansion ratio bell nozzle. Either way an E-D nozzle will still become under expanded once it hits it's critical pressure and the plume will grow.

Hmmm, I think your saying the E-D benefit occurs it lower altitudes and so is if no help here; that makes sense.

But the nozzle doesn't know if SABRE is in air-breathing or rocket mode, right? Maybe you simply meant earlier in flight when you said 'It's only an expansion deflection nozzle in airbreathing'?

The exhaust from the bypass burners (air-breathing mode) will affect the plume, but that's turned off by the time we're at the problematic speed & altitude/pressure. 

Not quite. The SABRE 4 cycle works by separating the rocket combustion chamber from the airbreathing combustion chamber allowing a high pressure rocket chamber (170 bar) and a low pressure airbreathing chamber ( 12 bar but can be from 6 bar to 20 bar depending on implementation). Each rocket  nozzle consists of a single rocket chamber surrounded  by 3 ( or more depending on implementation) airbreathing chambers, they share a rocket nozzle by having a 30:1 ratio inner nozzle after the rocket chamber throat and then a closable throat fed by the airbreathing chambers followed by a nozzle extension that takes the rocket expansion ratio to 120:1. The upshot being that in airbreathing mode the nozzle acts an E-D nozzle but in rocket mode the airbreathing throat closes and the rocket sees a conventional bell nozzle ( although they are designed to work concurrently to allow for a smooth transition).
My suggestion was to  move to a single large area ratio nozzle and use the airbreathing throat to do TAN as altitude compensation. My second suggestion is to crank the delta wing to increase the wingspan, my third is to consider alternate fuels as the SABRE 4 should allow the rocket mode to work on different fuels to the airbreathing mode and gelled hydrogen could be used for the airbreathing mode, together fuel volume could be slashed creating  a smaller tail.

Apparently I've not been paying attention because I was not aware that the SABRE 4 design - including this new dual-mode combustion chamber arrangement - had been made public! Was this image from a public document? Patent filing? If so, could you post links? Thanks.

Online pippin

  • Regular
  • Senior Member
  • *****
  • Posts: 2523
  • Liked: 276
  • Likes Given: 39
Re: The Reaction Engines Skylon Master Thread (5)
« Reply #544 on: 08/15/2015 05:55 PM »
Firstly the vast bulk of that exhaust will be in the plume. A very small part of it's mass flow will be hitting the aft fuselage.
If you mean that due to the fact that only a certain (angular) section of the overall plume hits the body, that's of course, true. What is this, like 10 degrees or so?
If you mean that the bulk of the exhaust is not in the shock front but heading straight aft you are wrong because it's just the nature of the over expanded flow that most of the actual plume mass will be in the shock front because it gets accelerated outwards by the pressure differential. Only a small part will be heading due aft and if my principles of conservation hold that part should not be relevant because at ambient pressure and probably temperature (see the charts) actually _all_ of the plume's thermal energy will be in the shock front.

And given the fact that there is pretty low ambient pressure and mass flow that overall thermal energy will be pretty close to the same as at the nozzle exit because there's not a lot up there the energy could be transferred to.

So I think it's pretty safe to assume that as long as the aft body is indeed within the shock front the thermal energy of the shock front hitting the body will be the same as the thermal energy at the same angular section of the nozzle exit. Which will be significant but should be somewhat easy to calculate if you know the details of the rocket engine.

Quote
Secondly this is a point simulation and Skylon is a on launch trajectory. It's outside air pressure is continually falling. That means the plume is growing but also the plume density is falling, lowering the energy of the gas that will hit the aft fuselage.
So where exactly is all that energy going in your assumption? I don't think you are right with this.

Quote
Being a point simulation you'd miss the integration of temperature involved. You've also missed that a large part of the body is at a much lower temperature and therefor that heat could be moved by a heat pipe arrangement. Such systems were looked at for the Shuttle wing but never flown.
Yes, that's true. I should have made a more precise statement: you can't just radiatively cool it in place unless you use something like the Shuttle TPS (note: I'm not saying the Shuttle's TPS could sustain it, I don't know the amount of heat transfer it could take, as you already mentioned we are talking about pretty low air pressure up there). Transferring the heat to other parts of the body and using these for radiation cooling or absorbing some of the heat using LH2 might work, I don't know.

I don't think something like the Shuttle TPS can be applied without changing the airframe design, BTW, because it will have completely different drag coefficients and will likely change the aerodynamics a lot.

Quote
Quote
This essentially means there's a significant chance the whole airframe concept is not feasible. It's the kind of risks you run into when designing all-new stuff and venturing into unknown unknowns with a lot of elements in your design.
Depends on your assessment of how much unknown you're dealing with.
Not my assessment but the report's.


Offline lkm

  • Full Member
  • ****
  • Posts: 508
  • Liked: 91
  • Likes Given: 1
Re: The Reaction Engines Skylon Master Thread (5)
« Reply #545 on: 08/15/2015 07:04 PM »
The use of altitude compensating Expansion/Deflection nozzles should help also.

It's only an expansion deflection nozzle in airbreathing mode, in rocket mode it's a 120:1 expansion ratio bell nozzle. Either way an E-D nozzle will still become under expanded once it hits it's critical pressure and the plume will grow.

Hmmm, I think your saying the E-D benefit occurs it lower altitudes and so is if no help here; that makes sense.

But the nozzle doesn't know if SABRE is in air-breathing or rocket mode, right? Maybe you simply meant earlier in flight when you said 'It's only an expansion deflection nozzle in airbreathing'?

The exhaust from the bypass burners (air-breathing mode) will affect the plume, but that's turned off by the time we're at the problematic speed & altitude/pressure. 

Not quite. The SABRE 4 cycle works by separating the rocket combustion chamber from the airbreathing combustion chamber allowing a high pressure rocket chamber (170 bar) and a low pressure airbreathing chamber ( 12 bar but can be from 6 bar to 20 bar depending on implementation). Each rocket  nozzle consists of a single rocket chamber surrounded  by 3 ( or more depending on implementation) airbreathing chambers, they share a rocket nozzle by having a 30:1 ratio inner nozzle after the rocket chamber throat and then a closable throat fed by the airbreathing chambers followed by a nozzle extension that takes the rocket expansion ratio to 120:1. The upshot being that in airbreathing mode the nozzle acts an E-D nozzle but in rocket mode the airbreathing throat closes and the rocket sees a conventional bell nozzle ( although they are designed to work concurrently to allow for a smooth transition).
My suggestion was to  move to a single large area ratio nozzle and use the airbreathing throat to do TAN as altitude compensation. My second suggestion is to crank the delta wing to increase the wingspan, my third is to consider alternate fuels as the SABRE 4 should allow the rocket mode to work on different fuels to the airbreathing mode and gelled hydrogen could be used for the airbreathing mode, together fuel volume could be slashed creating  a smaller tail.

Apparently I've not been paying attention because I was not aware that the SABRE 4 design - including this new dual-mode combustion chamber arrangement - had been made public! Was this image from a public document? Patent filing? If so, could you post links? Thanks.

Yeah, there were a whole bunch of patents made public in July, among them the heat exchanger frost control patent
(which SABRE 4 no longer uses), the SABRE 3 and SABRE 4  patents and the SABRE 4 nozzle patent. It's probably easiest if you just search back in the thread but this is SABRE 4.
It's interesting in that it is clearly Scimitar derived and the patent is broad enough to cover a SABRE variant with broadly Scimitar performance, save for the subsonic hub fan (see page 23, lines 16-22).


Offline john smith 19

  • Senior Member
  • *****
  • Posts: 5859
  • Everyplaceelse
  • Liked: 741
  • Likes Given: 4489
Re: The Reaction Engines Skylon Master Thread (5)
« Reply #546 on: 08/16/2015 11:35 AM »
If the work had been done in close collaboration with RE I'd have expected a response, or joint release. Alternately they were aware of it (and how it might affect investment) and this might be why there's been a softening of the response to TSTO suggestions?
And in a company with a better media strategy you'd be right. But REL don't really do media relations.  :(
Quote

It strikes me the plume is mostly going over the top of the craft. Does the simulation assume it will be following the same AoA in this steeper part of the ascent? Is that a reasonable assumption in the higher TWR and Mach phase of the flight?
I'd look a bit closer at those pictures.

I think most of them are plan forms.

BTW AIUI the comment in the 1st para of page 12 is wrong. The NASA calculated l/Cd ratio is betterthan the one REL conservatively estimated. The text does not match the numbers.

I don't think concluding that a TSTO vehicle is the way forward is really something NASA can say, although it sounds like something the USAF would like.

I'd like to see how these results stack up against the one REL got from DLR using their Tau code, which IIRC, does use real gas chemistry.

I will caution that both temperature and time are factors in wheather the vehicle can survive this sort of heating.

Checking the C1 spreadsheet shows that Skylon hits M6.41 45secs after going to full rocket power. It then accelerates to orbital speed over the next 240 secs.

So it has to survive partial immersion in this high temperature fluid for (at most) 4 minutes.

It would seem they need some system to extract the heat from the hot parts of the skin and move it to the colder parts so they can dump it.

Hmm.

Sounds like a kind of heat exchanger to me.

Wherever could REL find someone to design such equipment?

My suggestion was to  move to a single large area ratio nozzle
Judging by the diagram you supplied that's what already happens.
Quote
and use the airbreathing throat to do TAN as altitude compensation.
Assuming the REL E/D works already why would they need to do this? It puts Aerjet's IP in to their critical engine design path.
Quote
My second suggestion is to crank the delta wing to increase the wingspan,
Which gets what?
Quote
my third is to consider alternate fuels as the SABRE 4 should allow the rocket mode to work on different fuels to the airbreathing mode and gelled hydrogen could be used for the airbreathing mode, together fuel volume could be slashed creating  a smaller tail.
Except the SABRE 4 mode already cuts fuel volume needed. In one of their more recent progress reports to the IAS REL stated it's not the LH2 volume that's the problem it's the fuselage needs to be a certain shape and in fact there's a lot of empty  space inside it already.

REL have stated they will sub cool both LH2 and LO2 to below NBP to allow them to operate as no vent tanks on the runway.

You can call it "slush" or "gelled" Hydrogen but the fact remains its a 2 phase mixture of liquids and solids. It's something of a theme of REL's design work that they work very hard not to mix phases. In fact the whole point of the pre cooler is to cool but not to turn  it into a liquid. That is one of the enabling technologies of SABRE. I think the benefits for a 2 phase H2 flow would have to be very  compelling for REL to re-think this. Note that despite it being talked about since at least the mid 60's no vehicle has used it, just as (AFAIK) no engine has used gelled hydrocarbon fuels despite their (potentially) better safety and ability to carry energetic additives.
"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 lkm

  • Full Member
  • ****
  • Posts: 508
  • Liked: 91
  • Likes Given: 1
Re: The Reaction Engines Skylon Master Thread (5)
« Reply #547 on: 08/16/2015 03:53 PM »
If the work had been done in close collaboration with RE I'd have expected a response, or joint release. Alternately they were aware of it (and how it might affect investment) and this might be why there's been a softening of the response to TSTO suggestions?
And in a company with a better media strategy you'd be right. But REL don't really do media relations.  :(
Quote

It strikes me the plume is mostly going over the top of the craft. Does the simulation assume it will be following the same AoA in this steeper part of the ascent? Is that a reasonable assumption in the higher TWR and Mach phase of the flight?
I'd look a bit closer at those pictures.

I think most of them are plan forms.

BTW AIUI the comment in the 1st para of page 12 is wrong. The NASA calculated l/Cd ratio is betterthan the one REL conservatively estimated. The text does not match the numbers.

I don't think concluding that a TSTO vehicle is the way forward is really something NASA can say, although it sounds like something the USAF would like.

I'd like to see how these results stack up against the one REL got from DLR using their Tau code, which IIRC, does use real gas chemistry.

I will caution that both temperature and time are factors in wheather the vehicle can survive this sort of heating.

Checking the C1 spreadsheet shows that Skylon hits M6.41 45secs after going to full rocket power. It then accelerates to orbital speed over the next 240 secs.

So it has to survive partial immersion in this high temperature fluid for (at most) 4 minutes.

It would seem they need some system to extract the heat from the hot parts of the skin and move it to the colder parts so they can dump it.

Hmm.

Sounds like a kind of heat exchanger to me.

Wherever could REL find someone to design such equipment?
Active cooling would probably need hundreds of metres of their inconel tubing running underneath the aeroshell ceramic plus another hydrogen/helium heat exchanger and pump for the system. That's going to add a lot of mass. Hence the suggested mass trades to lower the surface area of the tail that needs it.

My suggestion was to  move to a single large area ratio nozzle
Judging by the diagram you supplied that's what already happens.
Nope, SABRE 4 still has 4 nozzles per nascelle, each with an exit area diameter of 1.4m and a expansion ratio of 120:1 for the rocket mode. My suggestion was that with just a  single nozzle  a diameter of up to 4.7m would fit in the same space and have an expansion ratio of up to 1000:1 or higher.  This would be a mass trade against how much active cooling masses but it would have the effect of shortening the time period when the nozzle is underexpanded dramatically and thus the heat soak that the tail experiences as well as lengthening the nacelle mitigating how much of the tail is exposed to the plume and how much the wing is exposed to  plume-induced flow separation.

and use the airbreathing throat to do TAN as altitude compensation.
Assuming the REL E/D works already why would they need to do this? It puts Aerjet's IP in to their critical engine design path.
REL's nozzle uses E/D for the airbreathing mode but the rocket mode is a pure bell nozzle so an extremely large expansion ratio version of it would need some form of compensation to prevent it being grossly under expanded in hypersonic flight,  my suggestion was purely based on the idea that the airbeathing throat already exists so it would only take a connection to the second fuel delivery system to us it for TAN delivery, I wasn't suggesting using Aerojets IP unless they own the very idea of it. However putting an E/D plug in the rocket nozzle might be an better solution.


My second suggestion is to crank the delta wing to increase the wingspan,
Which gets what?
It moves the nacelles further away from the fuselage trading wing mass for active cooling mass.

my third is to consider alternate fuels as the SABRE 4 should allow the rocket mode to work on different fuels to the airbreathing mode and gelled hydrogen could be used for the airbreathing mode, together fuel volume could be slashed creating  a smaller tail.
Except the SABRE 4 mode already cuts fuel volume needed. In one of their more recent progress reports to the IAS REL stated it's not the LH2 volume that's the problem it's the fuselage needs to be a certain shape and in fact there's a lot of empty  space inside it already.

REL have stated they will sub cool both LH2 and LO2 to below NBP to allow them to operate as no vent tanks on the runway.

You can call it "slush" or "gelled" Hydrogen but the fact remains its a 2 phase mixture of liquids and solids. It's something of a theme of REL's design work that they work very hard not to mix phases. In fact the whole point of the pre cooler is to cool but not to turn  it into a liquid. That is one of the enabling technologies of SABRE. I think the benefits for a 2 phase H2 flow would have to be very  compelling for REL to re-think this. Note that despite it being talked about since at least the mid 60's no vehicle has used it, just as (AFAIK) no engine has used gelled hydrocarbon fuels despite their (potentially) better safety and ability to carry energetic additives.
Slush hydrogen and Gelled hydrogen aren't the same things though, slush is a mush of hydrogen ice and liquid while gelled is a mixture of liquid hydrogen and a frozen second fuel, gelled is clearly easier to make and deal with and has safety and boiloff benefits beyond just fuel density.

The Skylon fuselage is shaped for minimum drag and mass, the report we're discussing suggests that their calculations haven't anticipated the drag correctly with regards rocket plume impingement and we've been discussing the mass effects of trying to mitigate such a heat soak, this suggests that maybe with a new design round taking all this into account the fuselage shape needs to change a little. As such maybe Skylon needs to be shorter.

The 60's argument is ironic considering we're talking about  a SABRE powered SSTO here. Slush hydrogen tankage and pumps were successfully built and tested for NASP and generally liquid ice is used industrially as a coolant so the technology for how to pump and tank slush is more mature than you'd think.

Offline adrianwyard

  • Full Member
  • ****
  • Posts: 947
  • Liked: 160
  • Likes Given: 209
Re: The Reaction Engines Skylon Master Thread (5)
« Reply #548 on: 08/16/2015 04:59 PM »
How critical is AoA during the latter part of the ascent? At some point the aerodynamic forces become negligible so might you be able to just aim the plume away from structure? (And retain control thanks to the center of mass & pressure being well managed in the Skylon config.) Perhaps even using negative AoA to keep it clear of the vertical stabilizer? Or, as it's a thin moving structure and cooling will be a challenge, move it to the nose - HOTOL style?

Offline Hankelow8

  • Full Member
  • *
  • Posts: 155
  • UK
  • Liked: 112
  • Likes Given: 40
Re: The Reaction Engines Skylon Master Thread (5)
« Reply #549 on: 08/17/2015 09:23 AM »
In 2012 The University of Strathclyde carried out a study for an alternate design for the Skylon airframe , CFASTT-1.
It concentrated on the re-entry heating characteristics of the airframe.

In view of the comments regarding rocket exhaust plume issues , it would be interesting to see if this alternate design improves rocket plume flow.


http://strathprints.strath.ac.uk/41933/1/Brown_et_al_Towards_Robust_Aero_Thermodynamic_Predictions_for_Re_Usable_Single_Stage_to_Orbit_Vehicles.pdf
« Last Edit: 08/17/2015 01:53 PM by Hankelow8 »

Offline adrianwyard

  • Full Member
  • ****
  • Posts: 947
  • Liked: 160
  • Likes Given: 209
Re: The Reaction Engines Skylon Master Thread (5)
« Reply #550 on: 08/17/2015 08:49 PM »
In 2012 The University of Strathclyde carried out a study for an alternate design for the Skylon airframe , CFASTT-1.
It concentrated on the re-entry heating characteristics of the airframe.

In view of the comments regarding rocket exhaust plume issues , it would be interesting to see if this alternate design improves rocket plume flow.


http://strathprints.strath.ac.uk/41933/1/Brown_et_al_Towards_Robust_Aero_Thermodynamic_Predictions_for_Re_Usable_Single_Stage_to_Orbit_Vehicles.pdf

Now you mention it, the CFASTT-1 had nacelle's that angle outwards away from the tail; presumably for aerodynamic reasons rather than heating. To my eye the new V-Tails are no closer to the plume than the single C-1 fin, but it's hard to be sure.

I noticed that the NASA Ames paper specifically states that nozzle gimbaling is not taken into account (p6) but that "the plumes are so under-expanded that it is unlikely this will substantially alleviate the impingement effects" (p15). If that's the case, then perhaps we can guess the CFASTT changes will also be of little benefit heating-wise.

And while we're talking of gimbaling, disregard my post from yesterday; I forgot that with Skylon the nozzles are already pointing down rather than straight back, away from the tail and structure.
 

Offline john smith 19

  • Senior Member
  • *****
  • Posts: 5859
  • Everyplaceelse
  • Liked: 741
  • Likes Given: 4489
Re: The Reaction Engines Skylon Master Thread (5)
« Reply #551 on: 08/17/2015 09:39 PM »
Active cooling would probably need hundreds of metres of their inconel tubing running underneath the aeroshell ceramic plus another hydrogen/helium heat exchanger and pump for the system. That's going to add a lot of mass.
A few A few hundred meters of this tubing will not weigh that much. bigger issues are likely to the thermal conductivity skin material.
Quote
Hence the suggested mass trades to lower the surface area of the tail that needs it.
Yes lowering the surface area is almost always the way to go.
Quote
Nope, SABRE 4 still has 4 nozzles per nascelle, each with an exit area diameter of 1.4m and a expansion ratio of 120:1 for the rocket mode. My suggestion was that with just a  single nozzle  a diameter of up to 4.7m would fit in the same space and have an expansion ratio of up to 1000:1 or higher.  This would be a mass trade against how much active cooling masses but it would have the effect of shortening the time period when the nozzle is underexpanded dramatically and thus the heat soak that the tail experiences as well as lengthening the nacelle mitigating how much of the tail is exposed to the plume and how much the wing is exposed to  plume-induced flow separation.
I think setting the expansion ratio on any launcher is tricky. Pretty much as soon Skylon hits the nozzles "preferred" altitude it will be passing through that pressure altitude.


and use the airbreathing throat to do TAN as altitude compensation.
my suggestion was purely based on the idea that the airbeathing throat already exists so it would only take a connection to the second fuel delivery system to us it for TAN delivery, I wasn't suggesting using Aerojets IP unless they own the very idea of it.
It's patented.  REL would have to license it.
Quote
It moves the nacelles further away from the fuselage trading wing mass for active cooling mass.
Sounds viable. The other option would be to widen the wings to put the final nozzle location as far back as possible. Again I think this needs a detailed look before major design changes are planned.
Quote
Slush hydrogen and Gelled hydrogen aren't the same things though, slush is a mush of hydrogen ice and liquid while gelled is a mixture of liquid hydrogen and a frozen second fuel, gelled is clearly easier to make and deal with and has safety and boiloff benefits beyond just fuel density.
It's still a 2 phase mixture and this is a design approach REL seem very keen to avoid.
Quote
The Skylon fuselage is shaped for minimum drag and mass, the report we're discussing suggests that their calculations haven't anticipated the drag correctly with regards rocket plume impingement and we've been discussing the mass effects of trying to mitigate such a heat soak, this suggests that maybe with a new design round taking all this into account the fuselage shape needs to change a little. As such maybe Skylon needs to be shorter.
Possibly, but the design trades on HTOL vehicles seem much harder than VTOL systems. What you gain on heat reduction you may loose too much control authority as the weight goes too far back, like HOTOL.
Quote
The 60's argument is ironic considering we're talking about  a SABRE powered SSTO here. Slush hydrogen tankage and pumps were successfully built and tested for NASP
Which one? The original NASP of the early 60's or the later X30 of the late 80's?
Quote
and generally liquid ice is used industrially as a coolant so the technology for how to pump and tank slush is more mature than you'd think.
Where LH2 with either SH2 or mixed with something like Methane I doubt anyone has long term experience on any kind of industrial scale.

Again AFAIK it's not the volume of hte propellant. It's putting it in a body that' big enough and trading the surface area. But there it's plum impingement area Vs propellant burn due to higher drag. [EDIT the different elements you can trade are complex, and it's very doubtful you can improve this item and not make any other factor worse, which I think is the difference between design IRL and design in a textbook.  :( ]
« Last Edit: 08/18/2015 02:56 AM by john smith 19 »
"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 lkm

  • Full Member
  • ****
  • Posts: 508
  • Liked: 91
  • Likes Given: 1
Re: The Reaction Engines Skylon Master Thread (5)
« Reply #552 on: 08/17/2015 11:39 PM »
Active cooling would probably need hundreds of metres of their inconel tubing running underneath the aeroshell ceramic plus another hydrogen/helium heat exchanger and pump for the system. That's going to add a lot of mass.
A few A few hundred meters of this tubing will not weigh that much. bigger issues are likely to the thermal conductivity skin material.
Well the Skylon D1 was designed to have 2.5mt of payload margin on top of the various built in structural margins so that's what they have to play with, and if they use it all they won't have anything left for the next problem. Perhaps they will need to re-reconsider the skin, Originally, in the very beginning, they planned to use C/SiC  for the Skylon skin which has a higher maximum temperature (1870K) and a lower density but they switched to System 2 as it was thought to be much cheaper to produce even though it increased the skin mass by 25%. Maybe they need C/SiC for the tail. 

Hence the suggested mass trades to lower the surface area of the tail that needs it.
Yes lowering the surface area is almost always the way to go.
Quote
Nope, SABRE 4 still has 4 nozzles per nascelle, each with an exit area diameter of 1.4m and a expansion ratio of 120:1 for the rocket mode. My suggestion was that with just a  single nozzle  a diameter of up to 4.7m would fit in the same space and have an expansion ratio of up to 1000:1 or higher.  This would be a mass trade against how much active cooling masses but it would have the effect of shortening the time period when the nozzle is underexpanded dramatically and thus the heat soak that the tail experiences as well as lengthening the nacelle mitigating how much of the tail is exposed to the plume and how much the wing is exposed to  plume-induced flow separation.
I think setting the expansion ratio on any launcher is tricky. Pretty much as soon Skylon hits the nozzles "preferred" altitude it will be passing through that pressure altitude.
Absolutely but if you could move the preferred altitude from say 30km to 90km for an acceptable mass/drag  increase then the time the tail sits in the engines plume would fall.  Skylon is , I think, pretty unique among rocket powered craft in having its engines ahead of any structure so while very large expansion ratios haven't previously been thought worthwhile, nobodies had to trade against extra TPS on their fuselage before. Having said that I'm actually doubtful of it mass wise.

and use the airbreathing throat to do TAN as altitude compensation.
my suggestion was purely based on the idea that the airbeathing throat already exists so it would only take a connection to the second fuel delivery system to us it for TAN delivery, I wasn't suggesting using Aerojets IP unless they own the very idea of it.
It's patented.  REL would have to license it.
[/quote]
I thought that their patents would be covering their scramjet derived injection process.

It moves the nacelles further away from the fuselage trading wing mass for active cooling mass.
Sounds viable. The other option would be to widen the wings to put the final nozzle location as far back as possible. Again I think this needs a detailed look before major design changes are planned.
It might also be an idea to angle the nacelles a bit like in the Cfastt-1 design.


Slush hydrogen and Gelled hydrogen aren't the same things though, slush is a mush of hydrogen ice and liquid while gelled is a mixture of liquid hydrogen and a frozen second fuel, gelled is clearly easier to make and deal with and has safety and boiloff benefits beyond just fuel density.
It's still a 2 phase mixture and this is a design approach REL seem very keen to avoid.
They were keen to avoid it in the substance their cooling because it meant they were wasting cooling potential, but they've never had a problem with it on the coolant side, after the hydrogen starts out in the system as a liquid before changing phase to a gas. 


The Skylon fuselage is shaped for minimum drag and mass, the report we're discussing suggests that their calculations haven't anticipated the drag correctly with regards rocket plume impingement and we've been discussing the mass effects of trying to mitigate such a heat soak, this suggests that maybe with a new design round taking all this into account the fuselage shape needs to change a little. As such maybe Skylon needs to be shorter.
Possibly, but the design trades on HTOL vehicles seem much harder than VTOL systems. What you gain on heat reduction you may loose too much control authority as the weight goes too far back, like HOTOL.

Bare in mind that if they need to make two thirds of the tail actively cooled then the tail will be noticeably gaining in weight and the cg will move back anyway.
 

The 60's argument is ironic considering we're talking about  a SABRE powered SSTO here. Slush hydrogen tankage and pumps were successfully built and tested for NASP
Which one? The original NASP of the early 60's or the later X30 of the late 80's?
For the x-30. I distinctly recall it from the book.


and generally liquid ice is used industrially as a coolant so the technology for how to pump and tank slush is more mature than you'd think.
Where LH2 with either SH2 or mixed with something like Methane I doubt anyone has long term experience on any kind of industrial scale.

Again AFAIK it's not the volume of hte propellant. It's putting it in a body that' big enough and trading the surface area. But there it's plum impingement area Vs propellant burn due to higher drag.

For Skylon it's probably not an ideal solution given the fuselage has to fit round a large payload bay  but I had originally been thinking about it for other SABRE use cases, say your the USAF and you have a 1mt 1m3 ISR payload you want to park over anyone anywhere at anytime, what's the smallest SABRE powered vehicle you could make. That kind of thing. For that I think a gelled hydrogen airbreathing mode and a subcooled propane rocket mode could make for a very much smaller vehicle.

I think the paper was pretty convincing that this is an issue but it was less so that they had adequately modelled the SABRE 4 engine and without a model of the actual heat flux the fuselage sees and how it reacts to it it's hard to know how much of an issue it is and if it is then the design trades are going to be very complex.

Offline flymetothemoon

  • Full Member
  • **
  • Posts: 239
  • Liked: 40
  • Likes Given: 212
Re: The Reaction Engines Skylon Master Thread (5)
« Reply #553 on: 08/18/2015 03:01 AM »
Could do with Hempsell popping his head in right now :)

Offline t43562

  • Full Member
  • *
  • Posts: 172
  • UK
  • Liked: 72
  • Likes Given: 36
Re: The Reaction Engines Skylon Master Thread (5)
« Reply #554 on: 08/18/2015 10:12 AM »
I wonder if Reaction Engines could do with their own CFD system so they can put in all the parameters they don't want to tell anyone and then see the actual situation?   

If anything needs to be done it would have to be checked again I would have thought.

... apparently there is this but I have no idea if it would really make sense to use it:
http://www.archer.ac.uk/casestudies/ara_casestudy.pdf

Perhaps it's the expertise that's needed that's the main problem?


Offline lkm

  • Full Member
  • ****
  • Posts: 508
  • Liked: 91
  • Likes Given: 1
Re: The Reaction Engines Skylon Master Thread (5)
« Reply #555 on: 08/18/2015 11:11 PM »
I wonder if Reaction Engines could do with their own CFD system so they can put in all the parameters they don't want to tell anyone and then see the actual situation?   

If anything needs to be done it would have to be checked again I would have thought.

... apparently there is this but I have no idea if it would really make sense to use it:
http://www.archer.ac.uk/casestudies/ara_casestudy.pdf

Perhaps it's the expertise that's needed that's the main problem?



It's probably worth bearing in mind that Reaction Engines today is a considerably larger company than they used to be, they're aiming to have 300 employees by the end of the year ( I believe) and have been hiring all sorts of experts over last two years. They undoubtedly have plenty  of engineers capable with CFD, they've certainly advertised for them, and they've previously commissioned the DLR to do CFD for them.
The upshot of all this is that Reaction Engines today is a whole lot more busy than they used to be and as such they haven't been presenting a lot of conference papers telling us exact what's what like they used to so while they're busy working on the Skylon D1 vehicle definition we largely have to rely on papers up to a decade old to tell us about Skylon construction. For all we know the D1 thermal environment has been properly modelled  and addressed structurally, internally and they just haven't presented a paper on it.
Anyway, with luck there will be some answers in October at the IAC when they present their next progress paper.

Offline t43562

  • Full Member
  • *
  • Posts: 172
  • UK
  • Liked: 72
  • Likes Given: 36
Re: The Reaction Engines Skylon Master Thread (5)
« Reply #556 on: 08/20/2015 08:39 AM »
... apparently there is this but I have no idea if it would really make sense to use it:
http://www.archer.ac.uk/casestudies/ara_casestudy.pdf

Perhaps it's the expertise that's needed that's the main problem?



....

It's probably worth bearing in mind that Reaction Engines today is a considerably larger company than they used to be, they're aiming to have 300 employees by the end of the year ( I believe) and have been hiring all sorts of experts over last two years. They undoubtedly have plenty  of engineers capable with CFD, they've certainly advertised for them, and they've previously commissioned the DLR to do CFD for them.
....
Anyway, with luck there will be some answers in October at the IAC when they present their next progress paper.

Undoubtely - I'm just basically wondering if they will need further help from Ames or if they can do what they need to do in the uk.  Ames has huge computing resources and if that was the limiting factor then experience wouldn't necessarily be able to make up for it.   On the other hand it might be some special code that only Ames knows how to deal with etc etc - I don't know.

Offline lkm

  • Full Member
  • ****
  • Posts: 508
  • Liked: 91
  • Likes Given: 1
Re: The Reaction Engines Skylon Master Thread (5)
« Reply #557 on: 08/20/2015 03:16 PM »
... apparently there is this but I have no idea if it would really make sense to use it:
http://www.archer.ac.uk/casestudies/ara_casestudy.pdf

Perhaps it's the expertise that's needed that's the main problem?



....

It's probably worth bearing in mind that Reaction Engines today is a considerably larger company than they used to be, they're aiming to have 300 employees by the end of the year ( I believe) and have been hiring all sorts of experts over last two years. They undoubtedly have plenty  of engineers capable with CFD, they've certainly advertised for them, and they've previously commissioned the DLR to do CFD for them.
....
Anyway, with luck there will be some answers in October at the IAC when they present their next progress paper.

Undoubtely - I'm just basically wondering if they will need further help from Ames or if they can do what they need to do in the uk.  Ames has huge computing resources and if that was the limiting factor then experience wouldn't necessarily be able to make up for it.   On the other hand it might be some special code that only Ames knows how to deal with etc etc - I don't know.

It seems REL has very little trouble gaining access to those sorts of research facility resources given the access to such resources they seem to have had already.
 I think that being located where they are at Culham Science Centre less than 30 mins from Harwell, they probably have more needed resources close by than any other location on the planet what with it being a nexus of so many high tech research facilities and companies enabling the skills and colaborations needed to be often within walking distance. Beyond that there really aren't very many hypersonic research progammes being funded today but there's a bunch of research facilities built to study the area and they'd all probably do research on an active programme that has a chance of flight.
Consider this, NASA published their 2015 draft technology roadmap in May,

http://www.nasa.gov/sites/default/files/atoms/files/2015_nasa_technology_roadmaps_ta_1_launch_propulsion_systems_final.pdf

The Roadmaps are a set of documents that consider a wide range of needed technologies and development pathways for the next 20 years (2015-2035) and under  Technology Area 1: Launch Propulsion Systems, TA 1.3: Air-Breathing Propulsion Systems it says

 "NASA is not currently advancing any technologies applicable for Earth-to-orbit in this area."

If you're a NASA centre with expertise in airbreathing and hypersonics, Skylon seems likely to be the only game in town for the next 20 years if you want to do any work on something that might fly.

Offline tatarana

  • Member
  • Posts: 10
  • Liked: 2
  • Likes Given: 0
Re: The Reaction Engines Skylon Master Thread (5)
« Reply #558 on: 08/20/2015 10:21 PM »
Regarding the cooling of the Skylon aft section, there could be a solution mentioned in a previous 2012 study by DLR, 28th  ICAS, attached to this post.
NUMERICAL INVESTIGATION ON THE POTENTIAL OF STEAM COOLING FOR THE SKYLON SPACEPLANE IN HYPERSONIC FLOW, by Th. Eggers

 It uses steam cooling, up to 100 [g/s m2], to achieve a sizeable reduction in local wing temperature.

Making a rough calculation, during 240 s (JS19 !), 20 m2 (wild guess), and 0,1 kg [g/s m2], it gives ~480 kg.
Remember much of this mass could be from water to be used for braking in case of abort, and will have been already launched (but not dumped off). The system to deliver it to the aft section and inserted in the air stream I can not estimate

It seems to me a high mass, but according to  <lkm> "the Skylon D1  was designed to have 2.5mt of payload margin on top of the various built in structural margins...".   

If it can solve this seemingly hard problem, why not ... ???

Offline john smith 19

  • Senior Member
  • *****
  • Posts: 5859
  • Everyplaceelse
  • Liked: 741
  • Likes Given: 4489
Re: The Reaction Engines Skylon Master Thread (5)
« Reply #559 on: 08/21/2015 08:43 AM »
Remember much of this mass could be from water to be used for braking in case of abort, and will have been already launched (but not dumped off). The system to deliver it to the aft section and inserted in the air stream I can not estimate

It seems to me a high mass, but according to  <lkm> "the Skylon D1  was designed to have 2.5mt of payload margin on top of the various built in structural margins...".   

If it can solve this seemingly hard problem, why not ... ???
Current plans call for dumping the water just after take off as successful take off implies it's no longer necessary.

However that was pre SABRE 4.

Sadly I think your figure on surface are is grossly out. This part of the fuselage is tapering (I think it's a Haak curve so a closed solution for surface area probably exists). But a 5m dia cylinder has about a 15.5m circumference and that section will be more than 1m long. For serious worst case numbers let's say a cylinder 10m long.  That's about 3720Kg of water. which is not good.  :(

Given 1 litre is 1Kg that requires a flow of 15.5litres a minute. For US readers that's about 240 US gallons a minute. With a 10 m/s flow rate that's a 45cm dia pipe. Provided valves are closed gradually water hammer (leading to pressures >3x delivery pressure) should not be a problem.

Clearly shortening the heating time and/or the surface area exposed pay big dividends.  :(

Before anyone gets too  concerned about those numbers please remember.

The rear fuselage is not a cylinder and the affected surface area should be substantially less than a cylinder.

The 240sec rocket flight is from the C1 spreadsheet and should also be substantially shorter for D1.

The figure for payload margin is payload. IOW the design reported in 2010 (provided it works) will deliver 17.5 tonnes, so would have to get a lot worse before it failed to deliver the stated payload of 15 tonnes.

IIRC the mass margin for the vehicle is 15% of empty weight.

BTW re-reading their 2010 progress report to IAC I noted that a goal of trajectory simulation was to keep temperatures during both ascent and descent at or below Shuttle levels.

While I doubt Skylon would fly the the Shuttle mix of tiles and blankets the fact that at no point does it exceed Shuttle skin temperature levels means such a design could use those materials if necessary.

The performance and maintenance cost hits would probably be dire but it does demonstrate that (in principal) a backup plan for the structure does exist and the design team is not "boxed in" if some unknown unknown surfaces with what they are planning to use.

It's little points like that that remind me that REL have spent a long time in studying the problem in  to the best resolution they could muster, and ensuring that where possible there is a backup plan. As funds have permitted they have increased that resolution in order to pick up more issues before construction begins, which is exactly the time you want to find them.  :(
« Last Edit: 08/21/2015 10:25 AM by john smith 19 »
"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.

Tags: