The Reaction Engines Skylon/SABRE Master Thread (6)

[ChrisWilson68]

John Smith 19 accused me of making a strawman argument.  In response, I listed several posts where he and others made the argument I was disputing, showing my argument wasn't a strawman argument.

Except the quotes you gave didn't make the argument you claimed they were making. I believe he strawman argument it was claimed you were making was not the pre-cooler is a critical component of SABRE, and SABRE is a critical component of Skylon, but that once the pre-cooler was 'solved' everything else would succeeded if given the specified money. I recognise that it wasn't explicitly stated to be the case, but you'd have to be trolling to insist it was about the pre-cooler.

So the first two quotes are about the pre-cooler being important to Skylon, which it is (or at least all versions of the engine so far. Frost control, it seems, was only important to the first three).

The first quote says the pre-cooler is "the unknown element" in SABRE.  The word "the" is even italicized.  The only reason to do that is to say it is the only unknown element.  That implies that everything else in SABRE is known, so given time and money, there isn't anything that would stop it from being built and work as expected.  Isn't that the point of making the distinction between "known" and "unknown" elements?

The second quote was in the context of whether REL has flown anything.  As a reply to the issue of whether they've flown something, it's clear the implication is that they've done the hard part and what's left isn't hard.  And that's for Skylon, not just SABRE.  Otherwise, it doesn't make any sense as a counter to the argument that REL haven't flown anything.

The third quote is slightly ambiguous - from context I'd say it was SABRE that "it" referred to - but even if "it" was Skylon the "if it performs as expected" as a qualifier to me quite clearly means it isn't taken for granted it would succeed give the required money.

It seems pretty clear "it" in the third quote refers to the pre-cooler.  So it's a claim that if the pre-cooler works to expectations, SSO will work.  SSO is a statement about the whole Skylon system, not just SABRE.

Even if "it" was actually supposed to mean SABRE, there's no escaping the conclusion that this quote is claiming nothing in Skylon outside SABRE could possibly be a showstopper.

The fourth quote is an opinion about REL's appraisal of their capability to build the engine, not Skylon. Even if it were about Skylon, it makes no claim as to REL's actual ability, only their assessment.

I claimed proponents of Skylon make this argument.  Here we have a claim that REL itself is making this argument.  Don't you think REL would count as proponents of Skylon?

The fifth quote, I think, is the closest to your characterisation, but the first half is again about pre-coolers and the second part says a lot of Sklyon isn't new - not that none of it is new.

The sixth quote would only back up your characterisation of the argument if you think John Smith 19 believes SpaceX were guaranteed to succeed, or than managing risk means eliminating it.

The fifth and sixth quotes were from the same post.  I tried to cut it down to representative pieces since it was so long.

John here claims that the pre-cooler is the only risky new technology in Skylon.  He lists a bunch of other elements of Skylon and for each claims it has already been done in some other vehicle.  Then he claims that given the funding, REL's chances of success are comparable to those of SpaceX when they started.

But SpaceX really wasn't doing anything new.  It was very clear when they started that there would be no technological showstoppers.  Given enough time and money, Falcon 1 would eventually fly to orbit -- nobody really doubted that.

[ChrisWilson68]

Perhaps we could bypass all this discussion of what people meant in past posts by just asking this question: Do you now agree with this statement: "The pre-cooler was the only significant risk of a technological showstopper for Skylon.  I consider the pre-cooler to have been proven and therefor there is not a substantial risk of a showstopper being found that makes Skylon impossible.  Given enough time and money, Skylon is highly likely to succeed."

If there aren't any people who agree with that statement, we can avoid a whole lot of arguing by agreeing that there are still significant technological unknowns for Skylon that could cause it to be infeasible.

[Paul451]

I was assuming they were drawing the line right on M5 while in the atmosphere.

So you made an assumption that made the idea as unrealistic as possible. You actually do that a lot when contrasting Skylon with any alternative, you seem to always pick the worst interpretation of any alternative. (Its actually really common in aerospace, not just amongst us space-nerds. Too many programs were destroyed because someone senior locked into a bad design early by false-comparing to any potential alternative, dooming the program. For example...)

[Re: MUSTARD] if 2 stages are not identical you more than double the budget

{sigh} No, you don't. Engineering never works that way.

That reasoning could only work if the biamese-stages are each of no greater complexity/cost than the non-biamese stages you are comparing to. And that assumption is always wrong.

Biamese/triamese designs never work as claimed. Identical designs for lower and upper stages never saves money compared to separate designs. The differences between the needs of the two flight-regimes always makes developing a single design that can perform both roles, more expensive than allowing the two stages to each follow their own logic. You can make some savings by sharing things like tank-section tooling, engine cores (for rockets, never for air-launched), transport, etc, but for a MUSTARD-style triamese the added complexity of the near-identical designs more than outweighs any savings in shared manufacturing.¹

Just as Skylon will be more expensive to design and develop than a conventional rocket stage - a SABRE-wielding first-stage will carry the overwhelming bulk of the development cost of a TSTO-system. Whereas the development cost of an expendable upper-stage will be on the same order of magnitude as the GTO stage that REL wants - ie, virtually nothing compared to Skylon-proper, and a small proportion of the development cost of a TSTO-Skylon.

Similarly the development cost of a recoverable/reusable upper-stage is primarily the re-entry system, which can be developed incrementally at fairly low cost once you have an commercially operational reusable first-stage.²

<sub>¹ The closest we've gotten is that the inefficiencies inherent in triple-coring the first-stage may not always outweigh the savings from common manufacturing, compared to just building unique side-boosters in their own factory. And likewise that the cost benefits of TSTO generally outweighs the theoretical efficiency of 3STO.

² The whole idea of expanding the payload market via a low-cost launcher is predicated on the idea that low cost launches also lower the development costs of novel in-space concepts, therefore increase the number of such projects. The same reasoning applies to the development of a reusable upper-stage once a reusable first-stage exists.</sub>

I can see a TSTO architecture where the staging Mach number is gradually raised, lowering the gross weight of the 2nd stage or increasing its payload up to the structural and thermal limits of the 1st stage carrier. [...]

While this is pedantically true, it's clear that when you're talking about staging at certain Mach numbers, you're still talking about in-atmosphere staging. What you wrote here is only true if staging is above the atmosphere, regardless of the staging velocity. Having variable super- and hypersonic staging is a terrible assumption.

It just seems like overkill once you've built a fully orbital engine to not use it as such.Looking deeper into it I just can't see the architecture.

Because you always assume that the first stage of a TSTO will always cost a similar amount to develop as Skylon. Therefore, to you it makes no sense to not develop straight to Skylon. And if your assumption was correct, that would be logical. But that assumption will always be wrong.

Forget whatever weird design you are obviously picturing for a TSTO. Just look at Skylon.

Can you understand that it is going to be structurally easier on the airframe, and structurally and thermally easier on the TPS, if Skylon was re-entering the atmosphere on a sub-orbital ballistic trajectory at Mach 6 or even Mach 10, compared to coming in at Mach 20+ orbital velocity? Can you understand that "easier on..." also means "easier to design"? And that "easier to design" means "cheaper to design"? And that "cheaper" means both sooner and vaster more likely to be funded.

At Mach 6, you are dealing with just 25% of the kinetic energy of Mach 20. At Mach 4, just 10%.

The only way for SSTO-Skylon to be cheaper to design and develop than Skylon-as-a-first-stage is if they've zeroed that difference. Not "reduced by", not "optimised for", but zeroed.

I know you hate the launch-vehicle-like-an-aircraft analogy but no one has ever built a large 2 stage cargo aircraft, despite the benefits of an enormous 1st stage to get a heavily loaded (and fueled) 2nd stage off the runway and airborne.

Actually it's quite routine.

Not only have a large number of experimental aircraft (Wright-flyer, first jets, many X-planes) launched via a separate system (catapult, sled, or jettisonable undercarriage, or launched off-the-back/under-the-wing of an existing aircraft), you also have expendable JATO units and drop-tanks used routinely by the military to boost take-off speed and range, respectively. And of course aerial refuelling is a standard for any mission where the requirements exceed the aircraft. Even when the aircraft is theoretically capable of the distance, the military still uses aerial refuelling or expendable drop tanks (or both!) to extend the performance (increased load-outs, increased loiter or combat time, higher speed for a given range, etc.)

Even with civilian operators, if the aircraft isn't capable of the range, they'll divide the route into stages and have refuelling at each stage. You don't think of that as a "multi-stage" aircraft because the same aircraft lands at each airport alone the route, but if airports along the route were impossible, if refuelling at those points was impossible, then obviously the industry would have had to copy the military model. Things like drop-tanks and aerial refuelling.

Same with multi-modal cargo. If you couldn't use locally operated trucks at shipping ports, the shipping industry would have had to ship its own trucks with every cargo. That drastically lowers your available cargo-mass per ship, which is why we don't do that, but if it was the only game in town, it's what we would do.

And flip side, if it was physically possible to build a spaceport at 50-100km, and ship fuel to that spaceport by less expensive means, it's obvious that we would never use multi-stage launchers, except for extremely unusual bespoke missions (as the military is to general aviation). [We might still have specialised vehicles for each stage - ground-to-magic-spaceport, magic-spaceport-to-orbit - because of the technological demands of each leg. Just as we use different vehicles for multi-modal transport - ships, trains, trucks, vans. But we'd never attach them to each other. If you suggested it, people would just look at you funny.]

The aircraft analogy only fails because people always use it wrong; just as you did.

[edit: minor]

[Paul451]

Perhaps we could bypass all this discussion of what people meant in past posts by just asking this question: Do you now agree with this statement: "The pre-cooler was the only significant risk of a technological showstopper for Skylon.  I consider the pre-cooler to have been proven and therefor there is not a substantial risk of a showstopper being found that makes Skylon impossible.  Given enough time and money, Skylon is highly likely to succeed."
If there aren't any people who agree with that statement, we can avoid a whole lot of arguing by agreeing that there are still significant technological unknowns for Skylon that could cause it to be infeasible.

I think there's another point that needs to made clearer to avoid some of the wasteful back-and-forth:

Do you agree that the Skylon airframe (ie, vehicle sans engine) represents the major cost of developing Skylon as a launcher?

Or instead do you believe that SABRE (or its successors) represents the main cost (or that the two, engine & airframe, are roughly equal)?

agreeing that there are still significant technological unknowns for Skylon that could cause it to be infeasible.

I think that people could agree with your statement, as written, even if they believe that Skylon is 90% engine development, 10% vehicle development.

[lkm]

Can you understand that it is going to be structurally easier on the airframe, and structurally and thermally easier on the TPS, if Skylon was re-entering the atmosphere on a sub-orbital ballistic trajectory at Mach 6 or even Mach 10, compared to coming in at Mach 20+ orbital velocity? Can you understand that "easier on..." also means "easier to design"? And that "easier to design" means "cheaper to design"? And that "cheaper" means both sooner and vaster more likely to be funded.

At Mach 6, you are dealing with just 25% of the kinetic energy of Mach 20. At Mach 4, just 10%.

The only way for SSTO-Skylon to be cheaper to design and develop than Skylon-as-a-first-stage is if they've zeroed that difference. Not "reduced by", not "optimised for", but zeroed.

This might be the source of much confusion. According to REL ( I believe Hempsell told us this on a previous thread) it's not actually clear that reentry is thermally worse on Skylon than Mach 5 air breathing as thanks to it's very low ballistic coefficient Skylon has a very benign reentry, at least according to DLR cfd modelling. One has slightly higher peak temperatures and the other longer heat sink, I forget which.
So at first pass any SABRE based TSTO flying the same sort of air breathing trajectory has to be equipped with the same level of TPS, at least until you have some modelling that says it doesn't need it. Also a TSTO is going to have a higher ballistic coefficient and be reentering lower so until you actually have some data I don't think you can actually confidently say that the reentry is going to be more benign that Skylon.

In short hypersonic air breathing is really hard on an aircraft, there's a reason REL believe Skyon is an easier build than A2, and whether you're going to orbit or not thermally and structurally it's likely  to be largely driven by that.

[t43562]

I note that this vacancy has been on offer since the end, I think, of July:

http://www.reactionengines.co.uk/careers_035markcommsmanager.html

I'm sure we're all looking forward to seeing it filled :-)

[Star One]

I note that this vacancy has been on offer since the end, I think, of July:

http://www.reactionengines.co.uk/careers_035markcommsmanager.html

I'm sure we're all looking forward to seeing it filled :-)

It closed yesterday surprised it's not been removed from the site.

[high road]

Considering this subject has been debated for five endless threads, might it not be time to give it a dedicated update thread, and (/or if it already exists) link to it in the first post of these semantics discussion threads?

Updates don't need to be major news, any news from a reliable source would be interesting.

[john smith 19]

This isn't really news but it's an interesting video about the testing of the ED nozzles:

Most interesting

The use of STOIC to investigate the air breathing to rocket transition is quite interesting since this was one of the other risk areas for SABRE. The fast acting flow valves that AE have developed are also quite intriguing. I wonder if they are a big enough improvement on COTS hardware AE might consider starting to sell them as a product?

I think the indication that the REL E/D nozzle design will be different and have lower cooling requirements is news, but it may have turned up before and I may simply have not have been following things closely enough to pick it up before.

[john smith 19]

Eh, no. Because it could be VERY likely that the prospective customer WANTS a TSTO powered by SABRE because he fully understands what he WANTs to get and those pushing SABRE as usable only on an SSTO airframe are unaware of, or just don't care what the CUSTOMER wants or needs. Saying it's illogical[/] is an attempt to 'intimidate' the argument into arriving at the preferred conclusion rather than actually using, you know, 'logic' and more importantly customer/market inputs which may in fact be the opposite of your own!

TBH I sincerely hope so.  . I just can't figure out where they want to draw the line that gives you a much better system.  The known  SoA for powered flight is the X15 but that was designed to operate long enough to soak the whole airframe.  I think you could use the same materials (since materials and structure would seem to be the the concern here) to build a SABRE powered 1st stage to go to the X15's top speed.  This would have major implications for the size and payload of any 2nd stage.

JS19 and others are simply making some assumptions that not even REL has made based on some assumptions that REL has made which pertain to operations and economics which themselves are based on some limited modeling for a specific outcome. And outcome which actually hasn't reflected the actual market or business planning needs of known users for several decades. REL knows this as they have been playing with words since the start to cover this point which is one of the reasons people have general doubts about how serious REL is and what, exactly, they are attempting to sell.

Care to expand on that? IMHO the economic model is every bit as important as the thermal cycle model.

Eh, again, that's not really accurate as the CoG and CoL shifts are well known and can be effectively countered with a number of configurations. I know you're aware of all the aerospace work on airframe design done over the decades and controllability of the various designs was only an issue with one specific rather recent design concept that started with some fundamental flawed assumptions. (And no we aren't talking NASP as aerodynamically it was fine, the main problems were unrealistic propulsion assumptions and requiring flying inside the atmosphere at Mach-25 in order to 'justify' afore mentioned propulsion system)

That's my point. People know this is an issue yet insist on designing vehicles that will have problems and then complain they can't get the design to work.

Any configuration which places its propulsion system near the CoG/CoL interface tends to negate the severity of the shift during high speed flight in a aerodynamic, lifting trajectory. Most combined cycle propulsion system vehicles are designed in this manner for that reason.

I think NASA was going to do one of these with the engines half way up the body.  Given that they've flagged plume impingement as a potential issue on Skylon I could not see how they'd avoid the rear fuselage getting quite "toasty."

I know this was mentioned but can't find it but that's the main reason the engines are on the wingtips because by placing them there you can in fact simply change certain assumptions and values and arrive at nearly the same outcome at the checkpoints along the flight trajectory without having to make significant changes to the overall "design" which has far to many 'down-stream' effects in any other position. "Skylon" is not an optimized design by any means and as e of pi points out the dry mass is optimistic to say the least, especially for a vehicle that is supposed to be able to be rapidly turned and serviced. Significant mass growth from design to operation is a given unless someone is willing to spend a LOT of money on materials and technology and that's not (as we're all well aware) very often conducive to economic construction or operation.

That's certainly been historically true. Manual tracking of part masses meant recalculating a new empty mass would be a major exercise.  But with BoM properties in a database (or at this level of fidelity a set of spreadsheets) propagating a part mass change through the design (although not the change on the mass of associated parts) should take seconds, giving an early warning that other part will need to be re-designed or re weighted. 

I seem to recall that the "Skylon" design can handle up to 15% lower engine performance and still be able to meet the given design goals. (I'll be corrected if I'm wrong, probably several hundred times I'm sure ) I thought I'd mentioned this but if so it probably got lost, but while that's a good margin it's actually not the one that people who actually design vehicles tend to be worried about. (A worry but not THE worry in other words)

REL treated Skylon  as aircraft, so they applied aircraft margins.

http://www.sciencedirect.com/science/article/pii/S0094576509004998

Interesting abstract.  The "exchange rate" of 1 Kg on the Apollo Ascent Stage --> 800Kg on the Saturn V demonstrates why Logsdon's and Africano's paper is well worth reading.

Ballpark inert mass growth during development can be anywhere from 10% (highly unlikely) to 50% (same) but is averages around 20% to 30% which also cuts into your margins.

With the proviso we're talking VTO TSTO rockets.

5% engine short-fall and 25% inert mass growth combined means "Skylon" has no payload, not to orbit anyway and there are standardized formulas for figuring a rough set of growth parameters for a vehicle design. I'm betting REL used them too. It doesn't really matter because aerospace designers and manufacturers use them too, for very preliminary designs and they rarely carry over to the actual vehicle.

REL have been very careful to work with industry standard methods so that anyone looking over their work (like potential investors) can follow the trail from objectives back to how they are going to be carried out. Sadly these include the cost models that predicted SX would have to have spent about $2Bn up to the first F9 flight, but they are the industry standard. 

And the bigger the vehicle the more likely for mass growth is. (Yes in some ways a bigger vehicle can absorb more and it averages out but as always ANY SSTO design is more sensitive to the combined factors. Less than a pure-rocket VTVL SSTO but it is not inconsequential. And yes a two or more stage design is also susceptible to mass growth but it actually starts with higher margins to begin with)

That ability to absorb mass growth would seem to be the virtue of those giant Philip Bono 1960's designs, although drag is a problem if you're going with the idea of a "micro launcher." so somewhere in between seems to be minimum trouble.

The main point of SSTO has always been based on the idea that a single-stage would be more economic to operate and in fact the amount of payload delivered, and where it is delivered is secondary. That turns out to be quite opposite of the requirements of the people who would be using them though. SSTO's, or pretty much any LEO orbital deliver systems, require to carry or be paired with on-orbit infrastructure, (usually in the form of some sort of Space-Tug or carried propulsion stage, Skylon's US, Shuttle Centaur, Shuttle Agene and PAM being examples) to meet market/customer requirements while, (obviously) multi-stage vehicles have this capability inherent. This is why straight up comparisons rarely work.

I'd point out that REL (as Hempsell said) sized Skylon to do a comm sat mission to orbit. Historically they assumed that the payload would carry it's own engine to handle GTO. Other payloads to LEO would then be free to use the payload for that stage for their own uses. Note it's the end users issue to get such a stage.  ESA advised them offering an upper stage would make the system more flexible, and now it's in the development plan.

Looking back it's a question of where you draw line on the stages maximum trajectory. Skylon's "first" stage gets you to orbit (which can be up to 600Km) , which for a lot of Earth Observation payloads would be enough. All other LV get you to maybe M10, but typically their 2nd stage can get you to GTO or even escape. A fairer comparison would be a TSTO just to LEO, then a third stage for anything about that.

A Skylon "first stage" is useful on its own. No other first stage is, unless people were prepared to do significant redesign and take a major cut in payload.

Your premise would seem to be fundamentally flawed. Nothing in JS19's post shows that SABRE is "stupid" to use in a TSTO design he simply STATES that it is by inferring it is ONLY capable of being used in that role with no supporting evidence.

Actually I checked my posts and have not called it stupid. I just could not see why you would. It seemed a "sub optimal"  to me. I can now see that an argument can be made but I've also recalled this came up with the question of how much a Skylon could carry sub orbitally. REL reckoned it was 2x their payload to orbit but the separation was quite risky and the payload would need something like 100-1500m/s (IIRC). Hempsell said they discontinued modelling on it as no one seemed that interested and you'd still need to build a Skylon to do it.

(He's also got an issue with inferring that it is the ONLY system capable of doing so which is also unsupported)

That's not quite correct. Pure rocket could do it if you can get the structural mass down, and that seems very tough. Anything needing deep pre cooling prior to REL's work I'm very doubtful about due to frost control and anything air breathing will need to have good air breathing Isp to swallow the loss in T/W. Anything that air breathing and  VTO means very low structural weight x (relatively) low T/W --> unlikely despite high AB Isp.

Nothing inherent in the SABRE prevents its use in a TSTO vehicle and REL has stated that it's not an optimum solution but have never given the impression that it's "stupid" or impossible. The main issue is that there is an assumption that a TSTO would not use the SABER in a similar manner to an SSTO for portions of the flight.

Which parts can SABRE not do?

Not stupid, just seemed an odd idea.

Further and probably more importantly Alan Bond (and this is NOT a dig or dissing him) is NOT actually a aerodynamics or spacecraft engineer, he's a mechanical and propulsion engineer. Lets be clear at least. No one at REL is in fact an experienced hypersonic aircraft designer or engineer (and there ARE a lot of those around) and the Skylon is "designed" to fulfill a set of criteria based on certain assumptions using the simplest possible vehicle design and standard DSMC modeling without that experience and expertise.

I'd be wary of that. Their work on E/D nozzle design (and improvement) and work on the inlets suggests they either have hypersonic experience on staff or access through partner companies.

Given the fact that Skylon as designed is only a very low resolution basic vehicle design for calculations of what is possible with the assumed performance of the SABRE engine the work done is sufficient to show that if the all the assumptions hold up the design is probably viable in basic function.

Which is pretty much SOP for all engineering projects, establishing that a design is possible if "Component X" can deliver these specs.

That does not translate to being viable or even desirable until a lot more 'variables' have been defined completely by people who have the applicable skill sets.

True, and a bit of a chicken and egg situation. Hopefully BAE will be able to provide some "sanity checking,"  although I would have expected the ESTECH  to identify any issues like this already.

When people propose something smart/reasonable alternative to skylon, like different aerodynamic configuration for SSTO craft using SABRE, they are treated well

Have you actually read and understood what that paper is about? First of all it's essentially attempting to use a greatly simplified aero-heating program in place of the standard, more complex one in regards to a "simple" aerodynamic shape as compared to a more "complex" one. (Bottom line is if you're going for very simple and quick calculations as long as you keep in mind the variables and limitations and there are a lot of them, the new program works to a degree) Secondly note that the configuration changes were to make "concessions to ease manufacturing and structural efficiency" but that it is essentially a re-skinned Skylon with more efficient aerodynamic design.

The paper doesn't show much new as REL was/is well aware that the Skylon is NOT an optimized design but a general one. It has issues which will require someone with more time, money and engineers to address sufficiently. And once having done so the result may (probably in fact) will look very little like the Skylon as currently designed.

Despite the look the cFASST-1 design changes very little of the basic Skylon design and already (this is shown both with the simpler HyFlow and the more complex and encompassing DSMC models) large changes are observed. It is obvious that more fundamental changes using well known high-speed/hypersonic methods will yield equally large changes and efficiencies.

For example:
Note the engines remain in the same place on cFASST. That is NOT because that is the only place that SABRE will work, nor is it because that position is the perfect position for an air breathing engine because it's very much not. Again it's to keep the calculations and formula simple so there is no need to calculate airframe/engine interaction. Experts are well aware that there are large increases in efficiency when engines and airframe are more integrated than when they are not, but the interaction also gets very complex very quickly. This is not in fact a bad thing as the more integrated the engines/airframe the greater efficiency overall both have. (For example both inlet and exhaust design gets simpler and overall propulsion efficiency increases when you can use the airframe as part of the system. It can cost some complexity in airframe thermal management but it can yield double digit efficiency % in overall performance)

I did not know it was that high.  But remember the old rule of software development
"Optimization is the root of most evil."

IIRC In the X30 programme "complexity in airframe thermal management" meant a)Building large chunks of design in RCC or b) going with active cooling.  AFAIK Skylon aims to make very limited use of either.

Everything has a price. Is that increased efficiency a) Affordable b) Needed ? NASP needed it with an engine  T/W ratio of 2:1 but does SABRE?

None of this precludes the design from being used in a TSTO launch system. The paper notes that "Optimization to include low-speed, low altitude flight will almost inevitably result in a configuration that is at least partially aircraft like and thus a vehicle which geometrically is more complex than space vehicles that have been designed in the past." That's very true but if you follow that logically then the conclusion also follows that something designed to be optimized for one regime will, by it's nature, NOT be optimized for different regimes. So compromises will be required at various points to allow a single airframe to cover all regimes.

The inverse problem is IMHO trickier. Ensuring any improvement in a subset of the flight envelope does not make other parts of the envelope worse to the point the design won't work. A TSTO would open up options at the cost of designing 2 vehicles to do so.

Whereas two vehicles both optimized for different regimes can, when used in combination be more efficient than a single vehicle.

Staging gets more efficient the higher and faster it's done. A SABRE powered lower stage, air breathing to @Mach-5 before switching to pure rocket can then accelerate outside the effective atmosphere to speeds approaching Mach-10 before releasing a space/very-high hypersonic optimized stage that carries the payload to LEO and beyond. Nothing in the SABRE design prevents this from being possible and it may in fact happen when the actual people who will be building the airframe get done with the trades and design studies.

With that staging velocity you could consider a very conventional expendable liquid upper stage. One notion I've wondered about would be wrapping 2 half circular wings rounds such a stage on pivoting on a structure like a piano hinge with some kind of "aerospike" or "aerodisk" to improve the aerodynamics of the flat front end. Such an empty stage should begin  entry at a very high altitude, but again has all the issues of being back end heavy of the shuttle without it's control surfaces.

[Rocket Science]

To jog your memory John, the X-15A2 needed repairs after the flight. The ablative TPS would need to be upgraded for a prolonged hypersonic flight regime...

[RanulfC]

True, but GTO is only part of the market and it's very challenging. Using the Skylon Upper Stage splits the design problem into the very tough part (Earth to LEO) and the relatively well understood process of going from LEO to GEO.

True, period actually While the following argument is technically true the main point being addressed is that by comparing and designing to the same abilities, (which is supposed to have been done with Skylon) as the EELVs the comparisons fully FAIL if they are only applicable to a part of the EELV mission profile and ability. GTO was in fact the MAIN mission of the EELV not delivery to LEO which arguably is the main mission of the Skylon design as that's what it's designed to do. GTO may be only "part" of the market but it's been the biggest segment monetarily and remains so and while the "tough" part may be surface to the point where the second stage is released to continue to GTO neither SSTO in general of Skylon in particular makes much difference. It's still "tough" and somewhere in there the second half of the journey has to be addressed and designed for and just because it's a relatively understood process doesn't make it simple or cheap.

Fair point  I was assuming they were drawing the line right on M5 while in the atmosphere. And I know what happens when you assume. If you want to go higher and above then SABRE would be the option. I was also thinking that this is already covered in the DARPA XS-1 programme.  I was also recalling  that REL are aware of the results of the cost modelling in the MUSTARD programme, specifically that if 2 stages are not identical you more than double the budget, because not only do you need their development and engineering budgets, you need one to cover the combined interactions of the stages.

The "line" of Mach-5 was given for the last Skylon trajectory I recall but somewhere between there and Mach-6/7 things begin to get 'difficult' if you want to continue air-breathing without significant design work and integration of vehicle/engines which REL appears to want to avoid. Somebody planning on doing it regularly, and economically will probably find the work more worthwhile since they are in fact doing a more in-depth and fuller analysis. The XS-1 program appears to be specifically avoiding any possible air breathing designs so is probably not a valid comparison. (It is also specifically aimed at vertical take off modes it seems) In general though air-breathing to Mach-5 (rarely up to Mach-8 without using a SCramjet) has been a general design cut off in concepts wanting to avoid excessive aero-heating issues no matter the engine placement or construction details.

It should be recalled that specifically the MUSTARD program was an attempt to significantly find methods of reducing the cost of a multi-stage launch vehicle WITH many assumptions loaded up-front that later turned out to be not as true as they were presented. The cost per stage modeling was one of those assumptions that proved less than accurate specifically as noted in one of the follow up posts. As to the interaction between stages, using MUSTARD for anything, specifically for something that's NOT a vertical take-off, pure rocket design, is highly misleading both in scope and results. Especially when you are required to use multiple vehicles of a non-optimized design as a basis.

The similar US IIRLV study, with initially similar assumptions but not tied to any single required design, noted that while identical stages would be optimum actually doing so would require much more expensive and complex vehicles and so recommended two different designs between the booster and orbiter for that very reason. The cost modeling showed that using two separate designs rather than a single non-optimum design was more cost effective overall. Hence the conclusions MUSTARD reached were in question already.

I can see a TSTO architecture where the staging Mach number is gradually raised, lowering the gross weight of the 2nd stage or increasing its payload up to the structural and thermal limits of the 1st stage carrier. But a reusable 2nd stage will still need the full orbital rated TPS anyway while an expendable 2nd stage will never give the cost per flight level of a fully reusable system.

The difference at the most basic level is that a designed from the start recoverable second stage can have optimized and cost effective TPS and structure which is now not required on the first stage. And is doing so can have more capability built in than an optimized SSTO design because of those capabilities. And careful with that second part as we already know "flight level" is a tricky measurement and expendables can in fact be cost effective under the right conditions

Lastly an optimized upper-stage can in fact have a much BETTER TPS than just "orbital" and can be optimized for much greater efficiency in that role which while it might be more expensive to design and develop will cost far less than something like the SUS or space tug in long term operational use.
 

That's fair also. It just seems like overkill once you've built a fully orbital engine to not use it as such. Looking deeper into it I just can't see the architecture.

You're not alone here as that's the view of every advocate of SSTO in any form I've ever talked to

The engine has always been only part of the equation and more often than not it's not even the biggest part thereof. And neither in fact is the structure, or method, or... Whatever. It has always been about the whole rather than individual parts that are used and quite often the most efficient use is non-intuitive, especially when there are basic bias' and assumptions involved.

Not going the whole way in a single stage suggests but you are using SABRE suggests you're OK with the engine but don't believe the structure.

Actually I'm agnostic about it and always have been, what I'm arguing is that assuming you know all the facts that are required 'solve' the equation before they are actually facts rather than assumptions, suppositions, and/or theory is a bit premature

The only large reusable high Mach structures I know are the Shuttle, the X37b, X15 and XB70. Except the X37b all have done powered flight inside the atmosphere but only the Shuttle did the whole potential SABRE speed range and only the XB70 could lift its own weight.

The X-37B actually does the whole range on the way back down, also Buran did the regime as well so you need to expand your examples Further "lifting its own weight" is a false argument, that's the engine not the airframe and the XB-70 did not in fact have a T/W of less than 1 at lift off. (You're also forgetting the SR-70 and B-58 both of which flew as fast as the XB-70 at times) You are trying to compare vehicles with vastly different designed missions and objectives as 'examples' of something which actually has nothing to do with the basic discussion.

Lastly how does it matter? None of the cited examples are or were optimized as multi-mach carrier vehicles with exo-atmospheric staging capabilities or single-stage-to-orbit vehicles so are you suggesting that because of those examples both Skylon and any type-of air-breathing, multi-stage vehicle are impossible? That'd be a rather funny argument don't you think?

Skylon is both bigger and heavier than the XB-70, it is meant to fly about as fast in the atmosphere as the X-15, (and around the speed at which that specifically designed vehicle was damaged severely) and into space and back bearing a payload less than the Shuttle's to a lower orbit. All in a single stage at greater economy than any possible other design. Or at least that seems to be the main point of contention

If you're that nervous only a rocket would be acceptably safe for the 2nd stage.

No idea where you got this particular 'issue' from or what it's based on but then again, so what? Why is a rocket powered seconds stage NOT acceptable? Oh and try to keep YOUR bias' and assumptions out of the equation where they are not specified or required ok? (See below)

It's this mix of optimism and caution that I'm having trouble with.

Yep and that IS your 'trouble' and not anything to do with the discussion

So an LH2 powered engine is OK.

An LH2 powered air-breathing engine capable of operation from zero to high-mach/low-hypersonic speeds? Yep. Can't see a single reason why they wouldn't be as they have been tested and studied since the mid-50s and shown to be workable and practical though no one has actually used one in operation yet.

A truss structure with fibre reinforced glass skin is not OK.

That's your assumption of something that wasn't said or indicated in any way, but specifically in my opinion it's perfectly acceptable IF that is where the actual, (and not simple) trades go. So far no one has actually gone beyond the simple trades.

Horizontal separation at high Mach number (that's an assumption but I've hear nothing about a VTO SABRE concept) is OK.

Actually I have in fact suggested that VTO SABRE is possible as you keep noting the T/W is sufficient for the job

Horizontal separation at "high-mach" numbers has in fact been done before, successfully. The B-58 was in fact designed around the concept. There are some design and operations factors to take into account but nothing that precludes the operation.

But lets be clear that any 'issue' of doing so is based on your assumptions rather than anything brought up in this discussion because that IS the very heart of your argument with the idea.

YOU keep assuming that (for some strange reason) any SABRE powered first stage must separate from its upper stage at high-mach INSIDE the atmosphere when the SABRE is perfectly capable of powering that same first stage on an exo-atmospheric path using the inherent rocket engines in it along with on-board oxidizer to achieve that trajectory.

Why? Because, I suspect, like most SSTO advocates it is difficult to imagine that if it's 'possible' to go from Mach-5 to Mach-20 using the same engine, airframe, and systems then why would you even imagine trying NOT to do so. The rather simple answer is because reality very often does not conform to what we might wish it to be but rather we most often have to conform to what reality is. Skylon's numbers are 'conservative' but are based on estimations and assumptions rather than actual numbers from research and construction data. They could be spot-on but the most conservative assumption would be they are in fact low by a good margin. Why? Because no one has built something like Skylon for operational use and neither REL nor anyone else can in fact truly state those at this point.

They have a lot of estimates that may or may not end up being accurate but at this point that is all they are and treating them as objective facts misuses the whole process.

True. I'd never really thought of Skylon's design being selected because it was easy to anlayse. It seemed to address quite a few problems.  I don't know about "optimum" but I'd certainly say "Good enough" and given the design goals I think it would be difficult to come up with something that looked much different but still gave undisturbed airflow to the engines and a well balanced design.

It's a cylinder with wings, with engines on the wing-tips to keep them out of the airframe flow equations. It is in fat about THE easiest theoretical shape to provide analysis on. Pretty much the same design Kelly Johnson started with it for what eventually became the SR-71 and for the same reasons. It 'solves' various problems by not addressing them therefore giving you a very basic idea of the requirements without resorting to far more complex, (and accurate) mathematical models. It gets you to "good-enough" for some basic design work but is about as close to an operational design as the CL-400 was to the SR-71 and for many of the same reasons. And I'll point out that the 'design goals' specifically are to have something that you can use to estimate the overall performance of the SABRE from using estimated performance of the SABRE itself. Everything from that point is estimates on top of assumptions rather than anything one could consider either a 'design' or a 'goal' in and of itself.

Lets be clear here, engine performance can be enhanced from 10% to 30% by optimizing the air-flow using the forward and aft airframe for pre-compression and expansion purposes. Controllability and stability can be highly improved by moving the engines from the wing-tips to positions closer to the airframe. Propellant feeding and flow would be far less complex with the engines within the airframe especially at high mach speeds. Airframe and wing stress' and manufacturing complexity are higher with the engines on the wing-tips as opposed to integrated within the airframe. There are high and low speed issues along with manufacturing and operational areas that could be addressed by a different, more integrated airframe and engine assembly.

Those are the most obvious 'issues' that the Skylon design ignores in order to simplify the equations and that is fine at this stage of the overall design process. But it is no where near a finished product and should not be treated as such.

AFAIK the perceived advantages are an easier design problem as your design is split in two and you can have more structure and a higher payload fraction.

But Skylon was designed to deliver an ELV payload fraction, not the 1% of the Shuttle. For an equal payload I find it very hard to believe a TSTO will be simpler to design, build or test.  I doubt it would be cheaper to operate either.  The big question is the perception of how risky is the Skylon structural design.

It's more than a perceived advantage and it's more than just about the delivered payload fraction. You may find it hard to believe but it is in fact something that is well understood in the aerospace industry. Yes if all factors are equal an SSTO would in fact be both economically and overall cheaper to build and operate but the plain truth is all factors are NOT equal. They never have been and while that may change at some future date, neither SABRE nor Skylon has enough actual performance facts to make the assumption they will do so.

Once SABRE has some actual performance data and Skylon is far better defined there might be reason to change, probably if either ends up coming in close to promises but then again maybe not as there are a lot of factors in play. And a vast majority of those 'factors' are at this point guess' and estimations because no one has yet built a fully reusable space launch vehicle of any type. A lot of those estimations are built upon rather shaky logic in fact as they tend to use known quantities from other forms of transportation as factors on which to base those estimations in ways that may not in fact be actually applicable to space transportation.

I know you hate the launch-vehicle-like-an-aircraft analogy but no one has ever built a large 2 stage cargo aircraft, despite the benefits of an enormous 1st stage to get a heavily loaded (and fueled) 2nd stage off the runway and airborne.

As Paul451 points out that's not really the case but more to the point the main reason I hate the analogy is because it's simply false from the basic premises. It is in fact constantly used wrong and the conclusions drawn are false and highly misleading.

NO TRANSPORTATION SYSTEM ON EARTH IS LIKE SPACEFLIGHT! Simply can't say that enough. No one has ever really NEEDED to build a "2-stage" aircraft (ship, train, automobile, whatever) because they could always find an intermediate place to 'stop' along the way. You can't do that between the surface and Earth orbit and the closest analogy is using multi-stage vehicles to reach Earth orbit. Full stop.

In fact lets take that too-often over used analog and spin it into the ground hard enough to leave a crater visible from orbit shall we Going from Earth to orbit and back using an aviation analogy is like taking off from New York and, no crossing the Atlantic isn't even close, crossing the Pacific non-stop is barely in the ball-park, no what you want to imagine is going all the way around the world, non-stop, non-refueled, and landing again in New York. (Oh and don't forget to kick the passengers/payload out of the plane half way around the world without landing because that would be cheating)

The simple fact is no one has EVER built a single-stage-to-destination long distance vehicle that operates on a regular, commercial basis in ANY form of transportation on Earth because we've never had any reason to do so. We've built some one-offs and some very expensive, limited use, government vehicles that are capable of some portion of that for specific reasons but we don't use them commercially because they are in fact not economic for any purpose other than the very specific one they are used for. No one builds them otherwise because we have never needed to.

It sucks but the simple fact is we HAVE to do so to access space because there IS no place to stop in between. And in doing so, if we ever can do so on a regular basis, will not only change the way we do things in space it will in fact require that we fundamentally change the way we do things in space. (Trust me that is WAY deeper than it looks and far to much to get into here )
 

If the design assumptions on Skylon are as conservative as you think that seems unlikely barring (again) the perception that HTOL SSTO  is risky and would essentially put the airframer at square 1 in design.

1) The 'design' assumptions on Skylon are quite conservative given the assumptions they are based on. Note they are based on assumptions from the start so 'conservative' is very relative in that context.

2) While not on square 1 any airframe developer is going to take into consideration what REL has done but they are going to be fully aware that unless every single item of the "design" of Skylon is absolutely on-target to several dozen decimal places the fact remains that no one at REL is in fact an airframe or aerodynamics specialist. So while they may take the same factors into consideration that REL used to arrive at the Skylon design but more likely they won't because they actually do know what they are doing and will not have the same bias' and assumptions involved that REL did.

3) You need to understand that ANY SSTO is "risky" due to numerous factors, some of which REL not only did not consider they specifically ignored in order to arrive at the Skylon "design" as it is. That's ok, they are not after all airframe designers and for what they need the Skylon "design" for they do not need to consider the overall vehicle in that kind of detail. On top of those there are manufacturing and operational details that REL is not qualified to detail because they in fact do not know all those details, nor do they pretend to know them. They have made some qualified assumptions in order to "close" the overall design in a method that allows them to carry out the calculations they need to carry out. They are not meant to be taken as literal and as far as I can see no actual manufacture or possible operator have taken them as such. Here seems to be the only place they are taken as gospel and for general discussion they actually perform satisfactorily but only as long as you keep in mind they are in fact assumed factors rather than actual ones and don't take them too far.

An interesting point. NASA pointed out that SABRE plume heating of the rear fuselage might be an issue, but did you have any others in mind?

Really nothing specific because the design isn't at a stage to get seriously picky yet. "This is a basic design of a vehicle that could possibly perform this mission with calculated thispayload using the SABRE engine with a performance of this assuming these parameters" pretty standard low-res data-sets and estimates overall. But in particular when someone with more expertise says there might be a problem it behooves you to take them seriously. There is room for tons of improvement from the current "design" which is to be expected at the current stage that "design" is in. But the mathematical models and calculations increase along the way which is something REL pretty obviously doesn't want to spend to much time or money on doing since they aren't an airframe designer. Once they have a partnership with an airframe designer who's working towards an actual vehicle design they will both put a lot more effort into working together to finalize and refine everything towards whatever design goals that eventually emerge from the trades.

Randy

[SICA Design]

Ranulf, I'll pick on a tiny bit of this novel you've written 

GTO was in fact the MAIN mission of the EELV not delivery to LEO which arguably is the main mission of the Skylon design as that's what it's designed to do.

...by re-quoting Hempsell from 'Thread 4 giving JS19 a jolly-good ticking off  :

...a whole range of mission requirements were considered and the one that drove all the sizing and mass constraints was launching a comsat on a cryogenic stage.  NOT because that was given a priority but because it needed the most, and everything else could live with less.  I am sorry John Smith 19 RanulfC you keep assuming comsats were the “primary mission” they weren’t, they were just the driving mission among a set of equals.

[RanulfC]

TBH I sincerely hope so.  . I just can't figure out where they want to draw the line that gives you a much better system.  The known  SoA for powered flight is the X15 but that was designed to operate long enough to soak the whole airframe.  I think you could use the same materials (since materials and structure would seem to be the the concern here) to build a SABRE powered 1st stage to go to the X15's top speed.  This would have major implications for the size and payload of any 2nd stage.

Actually the original construction of the X-15s was only for a bit above Mach-5 and as noted it was a different structure than anything anyone would consider using today mostly due to the expense and difficulty of maintaining that same structure over 200 flights. (Quick search didn't find it but I know it's been posted on the forums before. There was a very interesting study using the X-15 as a "comparable" vehicle for a recoverable first stage. It wasn't complimentary which was unexpected I understand )

Thing is it's right because the X-15 was specifically designed to test out high-speed aeroheating soaking but because there wasn't enough information, (duh, that's what the X-15 was for after all ) the design, materials and construction left a lot to be desired even in a test vehicle. And going past Mach-5.5 required re-building the basic airframe and the addition of ablative protection and even then there was damage.

The thing is we DID learn from the X-15 and we are well aware that it's design was limited, it's materials were not optimum, and it's construction was difficult to maintain and too prone to damage or failure under severe stress. Really we have much better designs, materials and construction methods if we want an effective, robust hypersonic vehicle. Getting someone to actually pay to build one now that's a different matter

But I'll point out that an effective hypersonic design is one that avoids the X-15, (Skylon for that matter) "fuselage and wings" design and the faster you go the more you want an integrated vehicle design overall. Mach-5 is pretty much the limit of that design which I suspect is one reason REL set that as a cut-off. Around Mach-6ish you want to have more body lift and the less "wing" area in the air-stream the better. That and combined low-speed and high speed aerodynamics is why you tend to see 'wide' body vehicles with small wings for hypersonic designs.

As hypersonic designs go the X-15 is terrible and the main baseline designs tend towards shapes like the X-24B/C, and NASP-ish/X-51-ish wide-body designs though there are alternatives even there.

Care to expand on that? IMHO the economic model is every bit as important as the thermal cycle model.

Not much to expand on really as the economic model is pretty simple overall except being based on an aircraft, (technically it would be more accurate to describe is as "any transportation system OTHER than spacecraft/launch vehicles" model but... ) model rather than a standard spacecraft/launch vehicle model. It has to be because none of the figures can be directly tied to actual data for the SABRE or Skylon because no such data exists. That doesn't mean you can't get some general numbers and as long as everyone keeps in mind the various assumptions in the models the basic output is good for a low-res, first-order outcome.

Probably the main sticking point though is it IS based a lot on aircraft economic modeling and like many other uses of "aircraft" as "spacecraft" there's some basic issues with using a lot of that data without qualification.

That's my point. People know this is an issue yet insist on designing vehicles that will have problems and then complain they can't get the design to work.

That's actually not a very valid point though because in fact no one does design vehicles with those problems because they know about them. HOTOL would be an example of that but that was BECAUSE the people designing it were unaware of some or all those issues. Did you have something else in mind because as far as I know that hasn't been an actual problem in any other vehicle design I'm aware of.
 

I think NASA was going to do one of these with the engines half way up the body.  Given that they've flagged plume impingement as a potential issue on Skylon I could not see how they'd avoid the rear fuselage getting quite "toasty."

They don't, no one does in anything past a first-order airframe design. The impingement is part of the engine exhaust system design and is dealt with by propellant circulation cooling along with fore-body pre-compression and airstream cooling in front of the inlets. It's not a problem if you design for it in the first place and the added efficiency is well worth the complexity over non-integrated engines. REL didn't bother with Skylon not because their set up is more efficient in any way, it's just easier to model the set up they have.

REL doesn't address impingement at all in the Skylon design because it would require a higher order of modeling which they don't need at the resolution they are using.

That's certainly been historically true. Manual tracking of part masses meant recalculating a new empty mass would be a major exercise.  But with BoM properties in a database (or at this level of fidelity a set of spreadsheets) propagating a part mass change through the design (although not the change on the mass of associated parts) should take seconds, giving an early warning that other part will need to be re-designed or re weighted.

No the properties data-bases themselves are low fidelity because each installation is different. Sometime only a little, sometimes a lot and changes in parts mass have a LOT of follow-on effects that low-order fidelity formula don't show. The fidelity is much higher once you have actual rather than inferred data to work with but at the moment low-order, low-fidelity numbers work for the design resolution involved. Nothing REL is using is more than a first-order fidelity at this point because they don't in fact have significant detailed data to work with, they have general number to plug into the data bases and get general results.
 

REL treated Skylon  as aircraft, so they applied aircraft margins.

Yes and that's not as "good" a thing as many seem to think. Skylon isn't an aircraft and many of the margins that apply to aircraft ONLY apply to Skylon in very limited and specific instances with many more equaling spacecraft and hypersonic aircraft margins. Neither of which is easy to calculate without direct data. Any spacecraft can in fact be treated as an 'aircraft' in some very basic formula work, it's a staple of most SSTO concept work for low-resolution concepts but once you move beyond the simple calculations you get more and more spacecraft data rather than aircraft. Aircraft data doesn't even work very well for economic models due to significant operational differences.

With the proviso we're talking VTO TSTO rockets.

Not really as part of the summery notes that it applies to, and from the aircraft industry as well. And specifically SSTO designs (both HT and VT) are mentioned as being effected. The paper is aimed specifically at mass growth in the Exploration Architecture but the general trending is historical over many different concepts and types. There is no indication that Skylon or SABRE would be immune from it and every reason to believe it will be effected.

REL have been very careful to work with industry standard methods so that anyone looking over their work (like potential investors) can follow the trail from objectives back to how they are going to be carried out. Sadly these include the cost models that predicted SX would have to have spent about $2Bn up to the first F9 flight, but they are the industry standard. 

So? The main issue is that what REL has done is not as high fidelity as is being assumed by many outsiders. Potential investors are not buying into Skylon, not directly because what REL has is not in any way to the point where it is realistic enough for investment on the level of actually building an airframe. It can't be because everything depends on the engines and until those are tested and actual data in-hand actual airframe design work can not proceed to a second-order concept and actual trade studies begin. How accurate or not their cost models for SX are in irrelevant, and actually they tend to cause conflicts over trivial issues as we're all aware.

Currently there is no real 'trail' for anyone to follow other than seeing how accurately REL has kept records of it's intakes and expenditures to reach this point in time. That gives some investors some confidence in RELs record keeping and budgetary process but it really doesn't begin to become relevant until they start full-up engine testing and begin defining a flight-capable/weight engine at which point they will actually have something to seriously interest investors. Skylon at this point is nothing more than a low-res concept with almost no relevance beyond some very simple calculation work on a possible end item to use the SABER engine.

That ability to absorb mass growth would seem to be the virtue of those giant Philip Bono 1960's designs, although drag is a problem if you're going with the idea of a "micro launcher." so somewhere in between seems to be minimum trouble.

It was touted as such but if you get into the details the problem was they HAD to be big and they still rapidly lost payload capability with mass over-runs or less efficiency than predicted. The problem is that in aerospace vehicle design smaller is ALWAYS cheaper BUT the devil is in the details The current spate of micro-launchers is pretty much based on industry and private reports of a high interest in nano/micro/small satellites by what's considered "non-standard" (non-government/non-sat-industry users) that could be a opening for enlarging the market. Significantly.

The main problem is, (other than it's pretty OT since I can't see Skylon or SABRE being downsized that much even though it would help in many, many ways ) that there's a good number of start-ups in progress and most of them are seeking not only the same market but a lot of the same funding. The fact that both government and "industry" have both announced and then canceled similar systems hasn't helped investor confidence either and that actually flows all the way to effect things like REL.

I'd point out that REL (as Hempsell said) sized Skylon to do a comm sat mission to orbit. Historically they assumed that the payload would carry it's own engine to handle GTO. Other payloads to LEO would then be free to use the payload for that stage for their own uses. Note it's the end users issue to get such a stage.  ESA advised them offering an upper stage would make the system more flexible, and now it's in the development plan.

And that was the initial mistake which as corrected by expert advice And anyone who really understood the nature of the Comm-Sat market would not have made that simple mistake because the payload NEVER provides it's own way to finish the job. That is specifically what the customer is paying the launcher to do, it's an issue the "end-user" neither wants nor technically needs to deal with. And not having that capability means you're not seriously interested in servicing that customer OR his market. Hence the ESA advice.

Hempsell or someone from the initial group should have know that but they were like most SSTO advocates which tend to have a significant blind spot to the "orbit-is-not-the-destination" problem with SSTO vehicles. LEO payloads are actually a bit more of a market than when REL first started up but that's not an excuse, that's a lucky break and it's still not the "comm-sat" market that was initially aimed for so it's a VERY big beginning mistake to have made. (No the SUS has not in fact made up for it either )

Looking back it's a question of where you draw line on the stages maximum trajectory. Skylon's "first" stage gets you to orbit (which can be up to 600Km) , which for a lot of Earth Observation payloads would be enough. All other LV get you to maybe M10, but typically their 2nd stage can get you to GTO or even escape. A fairer comparison would be a TSTO just to LEO, then a third stage for anything about that.

Eh no, looking back it's obvious that the initial concept of Skylon was not directed either at the supposed 'customer' it was advertised as serving nor is it doing so today and in order to correct that mistake a second stage, (one being designed outside REL which is both good and bad from a customers point of view) is being designed and built. You also can't 'compare' a TSTO just to LEO because Skylon still looses and you don't want to make the loss any worse than it already is. The specific TSTO that Skylon was designed against was an EELV which still delivers MORE payload "just" to LEO than Skylon and can then, without major changes or another stage deliver equivalent payload to GTO and beyond.

Changing the rules by insisting that adding a "third" stage to the TSTO simply shows how badly Hempsell and REL initially failed to address the market they said they were initially aiming for. An EELV can already do it and a well designed fully reusable TSTO could do it as well using only two, not three stages. Skylon can serve a limited market that has only become a viable market SINCE it was suggested and not even the one it was advertised as serving. In order to serve the market is was originally advertised to serve it requires two stages and not leaving that to the customer. That's something REL has to live with and so do Skylons advocates.

It's not a horrible sin mind you, but it's most certainly a failure to see and address the actual market and conditions from the start.

A Skylon "first stage" is useful on its own. No other first stage is, unless people were prepared to do significant redesign and take a major cut in payload.

First of all no it isn't. A Skylon "first-stage' has no real use as it doesn't exits yet. It has possible uses by itself if and when it has performance and manufacturing data and could, once in operation significantly change the way humanity access space. But it is only potential at the moment and has just as much chance of never going into operation at all. Meanwhile a functional and useful reusable first stage already exists with many of the same opportunities and more than a few drawback but in no way can you claim that no such stage exists.

Lets not get carried away with they hyperbole in the rush to advocate shall we?

Actually I checked my posts and have not called it stupid. I just could not see why you would. It seemed a "sub optimal"  to me. I can now see that an argument can be made but I've also recalled this came up with the question of how much a Skylon could carry sub orbitally. REL reckoned it was 2x their payload to orbit but the separation was quite risky and the payload would need something like 100-1500m/s (IIRC). Hempsell said they discontinued modelling on it as no one seemed that interested and you'd still need to build a Skylon to do it.

Well just because you, REL and Hempsell have lack of imagination... See the thing was while REL was making assumptions you needed Skylon, and the separation maneuver was risky, the simple fact was they never considered using anything BUT Skylon in the first place and their separation maneuver WAS risky because they were using Skylon, not because of the maneuver or process itself.

Stop trying to justify only using Skylon by only using Skylon and things are quite different, and the reason you would is quite plain actually. "2X times the payload of a standard Skylon flight to orbit" is a good start but as I recall the impulse needed was carried as part of the released mass but it was not if fact part of that "2X" payload mass which is part of the reason REL dropped it in the first place. Once you got past the strawman of using the Skylon itself for such a non-optimal use, (yes I very much agree that it wouldn't be what you want to use for first stage carrier vehicle, not odd at all as it's a very limited and point-design vehicle that has limited uses beyond it's primary job) and building in the staging maneuver without the restrictions and limitations that REL had in the scenario and you end up with much more 'bang' for your buck(s).

But again this all goes back to what Skylon was initially designed for and not any shortcoming with the SABRE itself.

 

That's not quite correct.

No it's actually quite correct because you have some serious blinders because of bias' and assumptions which I keep pointing out

Pure rocket could do it if you can get the structural mass down, and that seems very tough.

It IS tough and no SSTO advocate would argue that it isn't but it is far from impossible. Accept the limitations and you're fine. For one thing get over a fixation with LH2, (which has been difficult even for people who actually knew better but were blinded-by-the-light as it were ) and embrace alternatives you get much closer, much faster but there are limitations.

Anything needing deep pre cooling prior to REL's work I'm very doubtful about due to frost control and anything air breathing will need to have good air breathing Isp to swallow the loss in T/W.

Sorry more than one researcher working the "Aerospaceplane" in the mid-50s solved the frost problem and in more than one way as well. They considered it a non-issue and it's not their fault that REL didn't know about it the fact that it WAS solved was well documented. Air-breathing ISP and T/W are interchangeable by a number of well known and proven methods the big sticking point in most research was the "liquid-air" trap which along with SCramjets ate up all the research funding and effort for far to long. An air-breathing rocket was understood in the 1950s the only problem was the afore mentioned "liquid-air" side-track and while the actual deep-cooling effect was known it was lost in the data noise unfortunately. Patent's were taken out in the mid-90s in the US on a couple of deep-cooled air-breathing rockets using RL10s but no one ever was able to get funding to go beyond low-res concept work.

And that of course assumes that deep-cooling is really the end-all, be-all REL claims it is which is questionable as an assumption since there are methods of reaching Mach-5 without it and with pretty normal turbo-machinery. LOx injection makes a rather crappy rocket from an afterburner but it works well enough to allow exo-atmospheric staging possible if you don't want to carry a separate rocket system. In the end it comes down to trades as long as you're not too deeply invested in bias' and assumptions.

Anything that air breathing and  VTO means very low structural weight x (relatively) low T/W --> unlikely despite high AB Isp.

Depends really there were some pretty high T/W and isp engine systems developed. SERJ had a higher T/W and ISP than SABRE in LH2 systems (and no frost problems) and MUCH higher T/W (and middling but still SSTO possible) isp with Peroxide and jet fuel. VTVL TSTO was better in just about every aspect but like many concepts the designer got caught up in the liquid-air and SCramjet craze and separating out those aspects is more work than most current designers are willing to do for a system that was designed and tested almost 60 years ago.
 

Not stupid, just seemed an odd idea.

Not really and if you'd like to show exactly how any part of it is "odd" without resorting to requiring the Skylon as a vehicle feel free but I'm pretty sure without that last it won't seem as "odd" as it would seem

I'd be wary of that. Their work on E/D nozzle design (and improvement) and work on the inlets suggests they either have hypersonic experience on staff or access through partner companies.

Not at all really, it's the other way around. Again the Skylon as designed is the way it is to limit airframe airflow interaction with the nacelles and vice versa. This is so that limited and simple airflow calculations can be used on both the airframe and the nacelles. This requires work for calculation of the nozzle and inlet performance to be separate from that for the airframe which also requires very different and complex inlets and nozzles. Their work on improvement of both is because the Skylon design is not optimized and because they do not have the relevant skillsets and knowledge to do so. (Or don't want to which amounts to the same thing in general practice)

That they are getting some help from outside is obvious as they didn't significantly advance on the E/D nozzles until after RR joined up but so far we've seen nothing that wasn't already done along those line in previous work. Again this stuff is only to achieve clearer numbers on the engine performance OUTSIDE of any airframe interference which is REL's goal but not that of someone who is designing an air-breathing hypersonic airframe for a spacecraft.

REL has chosen to avoid as many problems as possible and a equal number of advantages in order to simplify their calculations on the operation of the SABRE by putting the engines where they are and designing the minimum airframe needed to have the needed figures for doing so. That in no way translates into it being the "right" answer at all.

Which is pretty much SOP for all engineering projects, establishing that a design is possible if "Component X" can deliver these specs.

Correct but it is never anything BUT a first-order concept at that point because "Component X" is plugged into the simple spreadsheets and formula but not actually integrated with all the necessary support and structural factors figured in. At this point you're only worried about the general placement and listed operation but you have not yet actually done any of the really hard work of integration and design. The design is "possible" only in general at this point without enough form and data to do more than estimate performance and capability. Which is fine but very far from something you can actually prove will do what you say it will.
 

True, and a bit of a chicken and egg situation. Hopefully BAE will be able to provide some "sanity checking,"  although I would have expected the ESTECH  to identify any issues like this already.

Many people assume that ESTECH did just that but what they ended up with is basically; "If all assumptions and calculations are valid then we see nothing in the given figures that will not work as planned" They wouldn't have actually followed up with in-depth research on each and every assumption or calculation to a degree much better than REL has done since that's not what they were asked. The "chicken/egg" at the heart of the whole concept is a working SABRE within parameters or at least nearly but no one can know how close everything else is until that data is available.
 

I did not know it was that high.  But remember the old rule of software development
"Optimization is the root of most evil."

That CAN be true of any type of project but you HAVE to make compromises or you end up falling short. Skylon is not optimized AT ALL and was purposely done that way so that REL would not have to do more work on the actual vehicle than they needed to arrive at some basic figures.

IIRC In the X30 programme "complexity in airframe thermal management" meant a)Building large chunks of design in RCC or b) going with active cooling.  AFAIK Skylon aims to make very limited use of either.

X30 is ALWAYS a bad example But, they were using RCC and active cooling because they were flying somewhere between Mach-8 and Mach-10 inside the atmosphere with NO other propulsion than air-breathing and they had to actually stay pretty low to make it work with such a small vehicle. The actual NASP was supposed to fly both higher AND faster but the X30 couldn't due to it's size. Active cooling is often ignored simple designs because it's seen as complex and a failure point. A bit true on both but Skylon actually uses it at well since the LH2 cools the wings as it flows to the engines and is circulated in the engines to keep the structure cool before being burned.

Skylon's not advertising it's use because it's a simple design but it's there anyway. Again the Skylon is designed to avoid airframe/engine interaction and not to be really good for air-breathing hypersonic flight even in a limited amount. It works but it's far from the best design.

Everything has a price. Is that increased efficiency a) Affordable b) Needed ? NASP needed it with an engine  T/W ratio of 2:1 but does SABRE?

Actually NASP started off with a T/W of something like 10+:1 at take off which went down as it ramped up to using SCramjets. Deep-cooling really boosts the T/W early in the flight. And IIRC no one ever admitted that the SCramjets were going to be less than at least "2 or more to 1" in any official papers despite that not being shown in any test data

Is it affordable? That depends on what the trades come up with but in general the current design of Skylon has NOT traded well over a more integrated design and REL is the only one using it. (Again that's for a very specific reason and one that has very little to do with performance and a lot to do with simplicity in many other areas) Does SABRE need it? Who knows since there is exactly zero actual data on SABRE performance. To put it mildly in the end SABRE DOES need it no matter what it's performance because any and all added performance only enhances it and opens up more possible benefits.

REL is actually causing some issue since they are so fixated with the nacelle/pod design and it's causing many people to wonder if the SABRE can in fact only be used in that configuration. ((Which I don't see as it does not in fact appear to have that issue to me though some folks I've talked to have brought it up)
Pod/nacelle designs have always had significant issues with use in hypersonic aerodynamics which is why most advanced designs do not use them. As far as I can tell most people who understand what REL is doing and why grasp the reasons and do not hold that against them but enough people have brought it up, along with other issues of Skylon as is being THE design, that I thought I should try and make it clear WHY REL has done what it's done and why Skylon is a good start but only a start.

In general I'm guessing I'm failing miserably

The inverse problem is IMHO trickier. Ensuring any improvement in a subset of the flight envelope does not make other parts of the envelope worse to the point the design won't work. A TSTO would open up options at the cost of designing 2 vehicles to do so.

Probably not as even at a higher cost it will probably be significantly easier and in the long run cheaper to optimize for each segment of the flight rather than the whole flight. Often the problem with SSTO is that due to optimization for the whole flight it IS certain segments of the flight where larger (and expensive) problems crop up. They key is designing smarter not harder but you have to really get into the design to find out which is which. Skylon isn't there yet by a long shot.
 

With that staging velocity you could consider a very conventional expendable liquid upper stage. One notion I've wondered about would be wrapping 2 half circular wings rounds such a stage on pivoting on a structure like a piano hinge with some kind of "aerospike" or "aerodisk" to improve the aerodynamics of the flat front end. Such an empty stage should begin  entry at a very high altitude, but again has all the issues of being back end heavy of the shuttle without it's control surfaces.

Well what kind of upper-stage to use is very much a function of mission and design trades but I was going to point out that one very 'nifty' side benefit of an optimized "high-hypersonic/space" stage could be the ability to perform plane changes with very little propellant using aerodynamic steering. Which in and of itself has all sorts of "multipliers" even if, (Ok especially if) you only use it with something like the 'nominal' Skylon. (Easier and more effective with a TSTO but I'll throw that out there anyway

Randy

[RanulfC]

Ranulf, I'll pick on a tiny bit of this novel you've written 

But what about the character development? The chase scene? Nothing about the unified field theory? Oh well, I tried

GTO was in fact the MAIN mission of the EELV not delivery to LEO which arguably is the main mission of the Skylon design as that's what it's designed to do.

...by re-quoting Hempsell from 'Thread 4 giving JS19 a jolly-good ticking off  :

...a whole range of mission requirements were considered and the one that drove all the sizing and mass constraints was launching a comsat on a cryogenic stage.  NOT because that was given a priority but because it needed the most, and everything else could live with less.  I am sorry John Smith 19 RanulfC you keep assuming comsats were the “primary mission” they weren’t, they were just the driving mission among a set of equals.

But that's NOT in fact what has been said constantly about what drove it AND while I understand the actual context of what drove the design along with what REL (and Mr. Hempsell) have been aiming at from the start, if that were the case they really should never have mentioned ELV and Comm-Sats at all.

I really DO understand what the 'equals' were that were being considered as they are obvious, SSTO was and remains the "main" force of what REL is trying to accomplish with SABRE if not Skylon but the rather threadbare fact that what is driving and has driven the design of Skylon and SABRE. To the point where REL has removed any mention of anything BUT SSTO operations in order to avoid the discussion. And while that simplifies things for them, (I'm sensing a pattern here ) it doesn't help that it avoids and does not address the actual issues of the design, the concept and their goals.

SSTO is in fact a goal worth striving for but as an engine manufacture and not an airframe designer from the start REL is going to have to face the fact they may not in fact get exactly what they want or deliver everything they think they can.

Thanks for reading though

Randy

[Hanelyp]

Lets be clear here, engine performance can be enhanced from 10% to 30% by optimizing the air-flow using the forward and aft airframe for pre-compression and expansion purposes. Controllability and stability can be highly improved by moving the engines from the wing-tips to positions closer to the airframe. Propellant feeding and flow would be far less complex with the engines within the airframe especially at high mach speeds. Airframe and wing stress' and manufacturing complexity are higher with the engines on the wing-tips as opposed to integrated within the airframe. There are high and low speed issues along with manufacturing and operational areas that could be addressed by a different, more integrated airframe and engine assembly.

Sounds like making the engines integrated with the airframe for airflow can improve practically everything except ease of analysis and aft-body heating.  Might even increase the speed the engines can practically breath air a couple Mach factors.  But first prove the engine works as well as projected, then worry about an optimal installation.

[Asteroza]

Lets be clear here, engine performance can be enhanced from 10% to 30% by optimizing the air-flow using the forward and aft airframe for pre-compression and expansion purposes. Controllability and stability can be highly improved by moving the engines from the wing-tips to positions closer to the airframe. Propellant feeding and flow would be far less complex with the engines within the airframe especially at high mach speeds. Airframe and wing stress' and manufacturing complexity are higher with the engines on the wing-tips as opposed to integrated within the airframe. There are high and low speed issues along with manufacturing and operational areas that could be addressed by a different, more integrated airframe and engine assembly.

Sounds like making the engines integrated with the airframe for airflow can improve practically everything except ease of analysis and aft-body heating.  Might even increase the speed the engines can practically breath air a couple Mach factors.  But first prove the engine works as well as projected, then worry about an optimal installation.

A good example of taking the spirit of skylon to one (of many) logical extremes would be sucking in the engines towards the center body and building the intake/nozzle ramps into the body sides (rather than the bottom), making something that looks like a flying axehead. Good illustration of that would SEI's Spiral-1/Sentinel design, in the following PDF on page 13-14.

http://www.sei.aero/eng/papers/uploads/archive/SEI_JANNAF_Sentinel_2007.pdf

Flying axehead is a simplified approach to getting the aero-integration, but forebody sizing would definitely cause it to mutate due to inlet/forebody interactions if one were to chase a cylindrical nose design.

Bad idea of the day, if using an axehead design with the typical low wing aligned with the bottom of the axehead, would you have a usable busemann supersonic biplane effect if you also had a second wing aligned with the top of the axehead?

[RanulfC]

Lets be clear here, engine performance can be enhanced from 10% to 30% by optimizing the air-flow using the forward and aft airframe for pre-compression and expansion purposes. Controllability and stability can be highly improved by moving the engines from the wing-tips to positions closer to the airframe. Propellant feeding and flow would be far less complex with the engines within the airframe especially at high mach speeds. Airframe and wing stress' and manufacturing complexity are higher with the engines on the wing-tips as opposed to integrated within the airframe. There are high and low speed issues along with manufacturing and operational areas that could be addressed by a different, more integrated airframe and engine assembly.

Sounds like making the engines integrated with the airframe for airflow can improve practically everything except ease of analysis and aft-body heating.  Might even increase the speed the engines can practically breath air a couple Mach factors.

Pretty much though as noted it makes figuring the engine performance alone really difficult which is why REL seems to have avoided it since the engine is in fact their main focus. As for the last part I have the impression that SABRE by itself has a pretty good Mach range (IIRC there was mention up to M-8 I think?) but the main question is how much you NEED to rather than what you can do since when you add in aerodynamic airframe heating at high hypersonic speeds it gets less viable without a huge amount of work. I think REL has struck a balance with picking the lower hypersonic end of things which many air-breathing advocates, (I'm looking at the SCramjet folks here ) overdo because they want to get the 'most' use out of the systems rather than looking at the entire mission objectively.

The long you use air-breathing the more complicated the whole vehicle gets not just the engines. And at it's most basic the SABRE and most other Combined Cycle Engines "solve" a different set of issues than just using atmosphere to power the engines. They get you from zero to hypersonic speed which for an air-breather and especially any horizontal lift design is huge advantage. Much beyond that, for space launch at any rate, you want to seriously consider getting up and out of the atmosphere which is where the Combined Cycles base on or incorporating rockets (such as SABRE) shine as they can do so while continuing to accelerate.

But first prove the engine works as well as projected, then worry about an optimal installation.

Yep

A good example of taking the spirit of Skylon to one (of many) logical extremes would be sucking in the engines towards the center body and building the intake/nozzle ramps into the body sides (rather than the bottom), making something that looks like a flying axehead. Good illustration of that would SEI's Spiral-1/Sentinel design, in the following PDF on page 13-14.

http://www.sei.aero/eng/papers/uploads/archive/SEI_JANNAF_Sentinel_2007.pdf

Flying axehead is a simplified approach to getting the aero-integration, but forebody sizing would definitely cause it to mutate due to inlet/forebody interactions if one were to chase a cylindrical nose design.

Nice, picked one of my favorites Just a couple of notes so we're clear; The Sentinel is a VTHL design so the wings can't support it for horizontal take off. Having said that you can see from the report that the take off T/W is rather crappy* and it's noted the design does "not benefit from augmentation below supersonic speed" which is exactly opposite of every ducted rocket, (which this engine starts out as) study results I've ever seen. Further the report actually calls out the design on the use of VTO with such a inefficient engine system.

Next the "axehead" (there's an official name for it somewhere I recall but that works for now ) design does not benefit from compression lift and external burning drag reduction in the exhaust stream like a design where the engines are located below the fuselage. At the same time it is far less sensitive to AoA to the airstream, but this goes both ways as a it does not benefit from positive AoA as other designs do.

But on the plus side it leaves both the top and bottom of the airframe more open and easier to incorporate things like landing gear and access systems without worrying about he engines, intakes, or exhausts.

*T/W at lift off for the study is 1.25:1 which is surprising as with an un-installed T/W of the engines being assumed between 20:1 to 35:1 for the type and specifically 27:1 for the study you'd assume a better total. Though SABRE is supposed to only have a T/W of 14:1 a big difference is that it has a much better integrated engine design than the "Independent Ramjet Stream/IRS," "Dual-Mode-SCramjet/DMSJ," and "Rocket" combined system base-lined in the study. This isn't surprising as the breakdown shows WAY the combined system is set up is directed to make the DMSJ the primary propulsion, (hello SCramjet research dollars) therefore the rockets have to be over-sized, (there's a development contract sent someone's way) and the "IRS" tacked on for... Well who knows really.

Problems with this set up abound with the DMSJ being only a small part of it, (for once not the main complaint ) but I'm still going to toss out that a SUB-sonic combustion ramjet will work up to Mach-8 so the actual justification for inclusion of ANY type of SCramjet is dubious to me. The "IRS" thing doesn't make sense to me as it appears to lack all the advantages of either an actual ramjet and an augmentation duct for a ducted rocket. The rocket system itself is very over-powered (despite the pathetic take off T/W it's bigger and more powerful than it needs to be BECAUSE of the VTO requirement) and while there are cases where physically removing the rocket system from an RBCC air stream have been argued, really it makes the design very much less efficient and the design of the ramjet much more difficult.

Maybe I should just point out that the AIRFRAME design is what I like not the concept itself?

Bad idea of the day, if using an axehead design with the typical low wing aligned with the bottom of the axehead, would you have a usable busemann supersonic biplane effect if you also had a second wing aligned with the top of the axehead?

Bad idea? Not really per se, the main question is what does it get you for whatever downsides there are in putting it into use With a Busemann Biplane the key to getting it to work well at more than just a single Mach number is having and keeping smooth airflow between the wings with a minimum, (preferably none) of disruptions to the flow. You'll note that most modern supersonic biplane designs have most of the "bulk" (fuselage, cargo, landing gear, engines sometimes though those are often between the wings as part of the separation structure but you will note they are set back from the leading edge of the wings) above or below the wings themselves to avoid their shockwaves.

To get the same effect with the axehead layout the wings would have to be pretty big. They would extend from in front to almost if not all the way to the aft of the vehicle and would include a number of electro-mechanical actuation systems to keep the air flow between the wings smooth. Make them "box" wings, (top one angled down, bottom one angled up and joined either by a vertical stabilizer surface or a structural joint) and you gain some low-speed advantages (more wing area with less actual wing, though how well this applies to a Busemann Biplane is questionable) with a stronger and more robust aero-surface area. Of course the wing-joint structures can be used to locate engines such as a couple of SABRE's so one wonders why you'd use the axehead airframe in that case and there is the basic question of how much compression you'd get from inside the wings and if the airframe and engine interaction would cause problems for the air-flow inside the wings. (My guess on those last two is, yes for the first but that that in and of itself would be highly problematical for the last)

To be honest I've heard of high-supersonic and hypersonic designs using something like "biplane" surfaces but they were in the context of using them to compress and act like high speed ramjets rather than actual lifting surfaces. So I don't really know if the idea would work, or even be practical in a general way. But I kind of like the idea for various possible designs. Now you've got me sketching out axehead airframes with paired trapezoidal (F-104 like) and delta box or joined wings, neat

Note: NASA/Glenn has a Turbine Based Combined Cycle simulator program here:
http://www.grc.nasa.gov/WWW/BGH/tbcc.html

I couldn't find one for Rocket Based Combined Cycle

Randy

[john smith 19]

I think REL has struck a balance with picking the lower hypersonic end of things which many air-breathing advocates, (I'm looking at the SCramjet folks here ) overdo because they want to get the 'most' use out of the systems rather than looking at the entire mission objectively.

In their 1989 Spaceflight article Bond said that was one of the key ideas behind SABRESkylon.  Not trying to hang on for just that last extra air breathing Mach number (and its associated friction heating) before shifting to rocket mode and going into a steep climb. It may also be the (rough) limit at which the design of the "spill ramjet," used to burn off the excess Hydrogen, starts to move into less grounded design areas. 

Maybe I should just point out that the AIRFRAME design is what I like not the concept itself?

When you wrote "This is one of my favourites" I was sort of wondering what part of it you favored. 

[RanulfC]

I think REL has struck a balance with picking the lower hypersonic end of things which many air-breathing advocates, (I'm looking at the SCramjet folks here ) overdo because they want to get the 'most' use out of the systems rather than looking at the entire mission objectively.

In their 1989 Spaceflight article Bond said that was one of the key ideas behind SABRESkylon.  Not trying to hang on for just that last extra air breathing Mach number (and its associated friction heating) before shifting to rocket mode and going into a steep climb. It may also be the (rough) limit at which the design of the "spill ramjet," used to burn off the excess Hydrogen, starts to move into less grounded design areas.

Wasn't the spill-ramjet dropped from the later designs? In any case there is a real issue of a loss of ramjet expertise over the last couple of decades from all the people and organizations that actually worked on physical rather than lab or theory ramjets. People who have done the research have shown that normal subsonic combustion ramjets are in fact a lot more capable than is generally believed, but the "common knowledge" today pretty much is that they don't work over Mach-5 and to go faster you have to have a SCramjet. Of course the "logic" from that point is if you're going to use a SCramjet you might as well get as much use out of it as you can, so...

The truth of course is different but between a lot of it still being classified, (which a lot of people don't understand since ramjets are 'obsolete' engines after all ) the general noise of the SCramjet advocates, and they die off of actual experienced people the knowledge is just vanishing at a frightening rate.

Maybe I should just point out that the AIRFRAME design is what I like not the concept itself?

When you wrote "This is one of my favourites" I was sort of wondering what part of it you favored. 

LoL,

Seriously for a non-body-compression lift design it's one of the better airframe concepts. (And it still actually manages to capture some good "body-lift" with the aft fuselage design when all is said and done)

It's actually a winged-body rather than the more often used lifting-body shape with minimal wings which is actually in some ways easier to manufacture and operate. (Something Skylon claims as well)

The wedge-shape gains much more capture area versus the overall vehicle wetted-area though it loses some body side compression in the process it's not as bad as most winged-bodies. (Skylon has far less but again it's base having as little interaction as possible from the start) This allows simpler variable compression inlets, probably not quite as simple as the under-body ramps but then again simpler than cone or nacelle's interacting with body and/or shock-on-lip compression. (Again Skylon avoids this by having the engines so far out on the wingtips, but in doing so also avoids all the advantages as well)

Overall it's got a better temperature distribution than normal winged-body or nacelle designs usually have which reduces the active cooling requirements significantly. (The Sentinel uses high-temp hydrocarbon in the form of JP-7 which is an operational assumption rather than an actual "requirement" so you can imagine how much better it would be with a cryo-propellant. And a non-REL related note the overall dimensions do NOT change if cryo-propane is used in place of the JP7 while they increase with methane and significantly so with LH2. Still I'd love to see a SABRE powered version done to the same fidelity. )

Things I don't like:
VTO. I understand the reasoning why, (mostly because it's an RBCC and Spiral 2, QuickSat as the TBCC was base-lined as the HTO vehicle which is biased in SO many ways from the start ) but it's really not needed for the overall design.

JP7 fuel. Again I understand why as the Air Force prefers "simple" hydrocarbon fuels but JP7 was designed and developed for the SR-71 and good lord JP10 is cheaper to use! If you have to use a hydrocarbon then seriously consider cryo-propane or methane. Heck I seem to recall a version of the QuickSat which used cryo-propane and no one batting an eye! In context I'd like to see this worked up with SABRE and LH2.

Partial exposure of the upper-stage/payload. Again I get the reasons but unlike the lifting-body vehicles there's a noticeable drag penalty for this in this design and fully burying it or fairing it over wouldn't have been that hard to fit in ONCE you got over some of the built-in flaws due to biased requirements.

Other than that, (and a few "peeves" here and there but that's more of those biased requirements than anything else) it actually is a favorite of mine for an air-breathing booster. The off-hand fact that you can, if you can ditch it intact that is, sail it into harbor like any other ship with a bow like that is totally coincidental I assure you

Randy