The Reaction Engines Skylon Master Thread (1)

[Robotbeat]

93143: The term you're looking for is: Extraordinary claims require extraordinary evidence.

No one is saying it can't possibly work. They just think it's pretty darned difficult and nowhere close to guaranteed. Skepticism is a rational response to breakthrough propulsion.

[93143]

93143: The term you're looking for is: Extraordinary claims require extraordinary evidence.

Which is either a pointless, overly specific statement or unscientific nonsense, as I have maintained before.  Why should the standard of evidence change based on one's preconceptions?  This is how people get locked into worldviews.

No one is saying it can't possibly work. They just think it's pretty darned difficult and nowhere close to guaranteed. Skepticism is a rational response to breakthrough propulsion.

No one has disputed any of that (except perhaps the definition of "skepticism").  The debate is about the specifics of how difficult and dubious the task is, and why.  I maintain that you and mmeijeri have been overly critical given the data.

[Warren Platts]

Of course expendable SSTO is possible. But it's less efficient from a payload-to-LEO:take-off-mass perspective (1% for SSTO, 4% for TSTO), so it's not desirable. And an SSTO of any sort can pretty much ONLY go to LEO, [etc.]

Dude, you're missing the entire point of the Skylon project. It most definitely is not an SSTO in the sense that you use the term.  It gets a bunch of it's launch mass through ISRU. Therefore, it's payload mass fraction is not 1% as you imply. It's more like 5%, better than your vaunted disposable TSTO. Honestly, I don't understand why you harbor so much FUD WRT Skylon when you don't bat an eye WRT SEP space tugboats the size of the Empire State Building.

[Robotbeat]

Of course expendable SSTO is possible. But it's less efficient from a payload-to-LEO:take-off-mass perspective (1% for SSTO, 4% for TSTO), so it's not desirable. And an SSTO of any sort can pretty much ONLY go to LEO, [etc.]

Dude, you're missing the entire point of the Skylon project. It most definitely is not an SSTO in the sense that you use the term.  It gets a bunch of it's launch mass through ISRU. Therefore, it's payload mass fraction is not 1% as you imply. It's more like 5%, better than your vaunted disposable TSTO. Honestly, I don't understand why you harbor so much FUD WRT Skylon when you don't bat an eye WRT SEP space tugboats the size of the Empire State Building.

Um, what? I was just using that as an example.

You're incredibly defensive, there, like you usually are whenever anyone expresses skepticism about concepts you like. I'm rooting for Skylon or any RLV. SEP tugs have no unproven breakthrough propulsion technology, just a scale-up of off-the-shelf (and the size of ISS or the several SIGINT birds would be perfectly fine as far as surface area). Not the same, not at all. And SEP tugs are not even close to a threat to Skylon (if anything, they make it worthwhile, since Skylon can't go beyond LEO).

Again, I'm rooting for ANY RLV. Definitely including Skylon. But guess what? You don't improve its chances by shouting down skepticism.

[john smith 19]

Of course expendable SSTO is possible.

His statement was *unqualified*. You're *assuming* that's what was meant. I'm sure  you know what happens when people assume things.

You might *think* SSTO possible ( and I certainly do) but the attitude I'm hearing is more along the lines of I'll only believe when it's already been done. I'd suggest that's far beyond "I'm not sure it can be done, and here's why."

I'll remind you again on assumptions about where an SSTO can go. part of where an SSTO can go is to do with weather or not it can be refueled in orbit and weather its engines can be re-started after a coast period. Both are options.

not so sure that's terribly realistic)

About as accurate as Spacex's reusable LV animations in some ways.

and very difficult. If the pre-cooler gets clogged with ice (hard to see why it wouldn't, although they certainly assert it won't),

The dirty little secret of *every* airbreathing systems.

It's why one of their first test programmes was to test their "frost control system." It's the one part of the design they will not discuss at all but which was demonstrated to the ESA evaluation team.
It's mentioned on both their web site and the 1989 Spaceflight article.

William Escher and Paul Kryze (major supporters of airbreathing in the US) have stated this is a key issue (and it's one the available reports on the "Aerospaceplane" programme in the 1960s never resolved)

The design is non conservative where it cannot *afford* to be conservative.

<long series of materials and design points snipped>
All true in 1950, before the Suntan programme developed what would become the drive turbine for the RL10. Or 1960, before the J2 demonstrated that large LH2 combustion engines could work and combustion instability was controllable. Or 1970, before the SSME showed they could work at chamber pressures of 200atm (BTW there was a DLR programme in Germany in the late 60's that also did LO2/LH2 staged combustion to > 200atm)  and above. Or 1985, before EADS/Astrium showed you could engineer a large LO2/LH2 in Europe as well.

I'd suggest it's no accident that at least some of the parameters of SABRE match that of Vulcain or Vulcain 2 on the Ariane.

There is a *huge* amount of materials properties information available. How much REL could use for legal reasons (ITAR specifically) is open to debate but that materials database *exists*. Even the visco-elastic-plastic behavior materials model developed to explain why the SSME combustion chamber life was 1/3 what the simple (or rather simplistic) models in use at the time predicted it would be, along with the guidlines of when they are still valid. BTW Friction Stir Welding was invented in the UK by TWI, who also worked on Shuttle tile issues (probably their bonding to the fuselage). Sklon baselines a regular Aluminum alloy for the tanks but AFAIK Ariane is now made using AlLi, as are substantial parts of the Airbus A380. The 1st gives cryogenic experience, the 2nd gives production volume.

The fuselage structure is more speculative and I'd expect the materials properties DB there is more proprietary to REL.

Managers use Powerpoint. Engineers use spreadsheets, CFD codes or write their own software for this.

REL closes both the engineering *and* the business case. That's why they are getting funding and others are not. When funders plug REL's numbers into their models they get the *same* answers.

Very roughly the "conservative" 2 stages designs come out at 2x the cost of 1x. REL staff worked on the BAC "MUSTARD" project cost modelling at how to develop a trimese LV for the cost of a single stage. The answer. It was damm hard. They also had 1st hand experience of Concorde, where there where it took 3 attempts to get the partners to agree a minimum passengers/maximum range design as the best option.

RLV *will* be >1x the cost of ELV by a factor dependent on the proposers optimism and awareness of engineering reality.
The present ELV business model leaves all the *operations* cost with the mfg (or it's subsidiary)

Result TSTO development cost + # of launches (from the launch market model) * ops costs >> operating profit from project launch market, unless developer gets whopping "Assured access to space" payment from the govt.

Conservatism and "incremental" improvements will simply *not* cut it. point solutions that plan to build (and operate) a few vehicles because there's an old Cold War supply of super performance engines will not cut it either, not when you factor in the Mach range, novelty, staging issues etc. What you gain on the engine cost you loose in the novelty of the development programme. Either it assumes no surprises or it's costs exceed *plausible* profits once the operations costs are factored in. It's discounted rate of return on investment is -ve. Game over.

You either cut the number of stages or get the users to take on the operating costs or both. IOW *sell* it to them and make it simple enough they don't need to hire a standing army to support it between uses.

 It's the UL about the Bell system before Strowger predicting that by 1918 (or thereabouts) that to meet demand *every* US citizen would have to be a telephone operator, which in a sense is *exactly* what happened.

I'll emphasize this is *high* risk / high reward *development* programme.

It's as certain of success as the X1, X15, Atlas pressurized tank precursor vehicles or STS *before* their 1st flights.

Anyone thinking it's just a case of placing the orders for the materials and machinery is deluding themselves. But investors don't mind risk if the developers look like they have a *realistic* understanding of those risks and the *reward* matches it. That's why it's called "venture" capital.

[mmeijeri]

His statement was *unqualified*. You're *assuming* that's what was meant. I'm sure  you know what happens when people assume things.

It is indeed what I meant, and actually I did qualify my statement, though maybe not in every sentence.

[mmeijeri]

William Escher and Paul Kryze (major supporters of airbreathing in the US) have stated this is a key issue (and it's one the available reports on the "Aerospaceplane" programme in the 1960s never resolved)

I've heard of Escher, but never of Kryze. Don't you mean Czysz?

[mmeijeri]

The design is non conservative where it cannot *afford* to be conservative.

Where it cannot afford to be if RE wants to do SSTO. Yes, they don't take unnecessary risks given their goal, but SSTO is a self-imposed constraint. They may believe it is economically necessary if we are to have a breakthrough in launch costs (and prices), and they may even be right. But it is too early to know this for a fact. Just as it's too early to tell if SSTO RLVs are even possible with all-chemical propulsion. There is more than one reasonable approach to cheap lift.

Very roughly the "conservative" 2 stages designs come out at 2x the cost of 1x.

Perhaps, but a reusable HTHL first stage with an expendable upper stage would be much cheaper to develop. It would not reduce specific launch costs by as much of course, but it's still a plausible approach. And it's certainly easier.

Conservatism and "incremental" improvements will simply *not* cut it.

Unsubstantiated. And so far large, highly ambitious projects have not cut it either.

[mmeijeri]

I'd suggest it's no accident that at least some of the parameters of SABRE match that of Vulcain or Vulcain 2 on the Ariane.

It seems reasonable enough to assume the rocket part of SABRE can be made more or less as RE intend to do. But remember that even the much less ambitious Vinci has made very, very slow progress and that we still don't have ESC-B.

[Rugoz]

So what are the disadvantages of using a skylon type vehicle as a reusable booster/first stage in the case the SSTO skylon is too risky for investors?

Some that come to my layman's mind:

- Shielding the the upper stage from the aerodynamic heating at mach 5.5 (Conventional rockets leave the atmosphere at lower speeds, e.g. Ariane 5 makes approx. half of mach 5.5 at an altitude of 25km).

Although the upper stage could be released at a higher altitude (suborbital trajectory) and be placed in a big cargo bay.

- Horizontal transport of the fully fuelled upper stage requires increased structural strength.

- And of course less reusability, higher production costs.

What to the experts think? 

[flymetothemoon]

They may believe it is economically necessary if we are to have a breakthrough in launch costs (and prices), and they may even be right.

Surely that should be the goal? To not be a sheep and achieve something truly remarkable? ESPECIALLY if the costs are similar to pursuing something uselessly conventional. There is no in between. In between maintains masses of the cost and complexity of development (+production) AND operations that Skylon eliminates. In between will see us having these interminable conversations for another 50 years.

Very roughly the "conservative" 2 stages designs come out at 2x the cost of 1x.

Really? For development, production, operations AND turnaround time?

Unsubstantiated. And so far large, highly ambitious projects have not cut it either.

If you try to make a turkey - even a large, highly ambitious turkey - AND you succeed, what you have achieved is a turkey. A large, ambitious turkey.

I love hearing some of the things Elon Musk says. I'll paraphrase, with apologies, "If the thing you are planning to do is intrinsically dumb and pointless, erm, why are you bothering?". He says they call fuel cells "Fool Cells" at Tesla bacause the figures stack up so badly against batteries.

Why would you think of doing e.g. NELS if you already have a good idea that Mr Musk will succeed wth his re-usable system? Especially when his expendable system is probably already far cheaper than NELS could ever be anyway.

[mmeijeri]

Surely that should be the goal? To not be a sheep and achieve something truly remarkable?

Absolutely, and there are conflicting theories about the best way to go about it.

Here's a link to that paper I mentioned earlier:

Conflicting Heuristics for Low-Cost Launch Vehicle Architectures

ESPECIALLY if the costs are similar to pursuing something uselessly conventional.

That's a big if...

There is no in between. In between maintains masses of the cost and complexity of development (+production) AND operations that Skylon eliminates. In between will see us having these interminable conversations for another 50 years.

It's possible, but there's no way to know just yet. There are many plausible RLV concepts that could dramatically reduce launch costs. SpaceX is working on one as we speak, as are Blue Origin and several smaller players. So far all of them are ahead of RE, though that is no guarantee they will stay ahead or that they will ever achieve orbit with an RLV.

Why would you think of doing e.g. NELS if you already have a good idea that Mr Musk will succeed wth his re-usable system?

What is NELS?

[Robotbeat]

...ESPECIALLY if the costs are similar to pursuing something uselessly conventional....

Skylon is /the/ most expensive RLV project out there (much more than a conventional expendable, for sure), as far as estimated/likely costs for development. Blue Origin and SpaceX's vehicles are likely to have development costs an order of magnitude less, for roughly the same payloads (Blue Origin's will likely be less, but not uselessly so). Lockheed Martin's reusable booster will be less expensive for development. We don't know enough about XCOR's project to say for sure, but it's certain to have lower technological requirements, which usually translates to lower development costs.

What's better? A single, grand project with tighter constraints and higher risks or half a dozen smaller projects that each have lower risks and a few may even have lower per-launch costs? It remains to be seen, but I subscribe to the school of thought that pursuing multiple independent projects is very often better than a single big project, at least on this scale. I hope Skylon does manage to get full funding from Europe so we can find out.

[john smith 19]

I've heard of Escher, but never of Kryze. Don't you mean Czysz?

I do. Doing it for memory can be risky.

[flymetothemoon]

ESPECIALLY if the costs are similar to pursuing something uselessly conventional.

That's a big if...
[/quote]

Certainly not an order of magnitude different according to Reaction Engines. They are still talking roughly 'double' NELS at the moment which is due to lift roughly half Skylon's capacity. Although we haven't heard the revised costs for Skylon D1 and likely won't for some time. No point in coming up with a figure I guess until (we hope) engine development is complete.

Why would you think of doing e.g. NELS if you already have a good idea that Mr Musk will succeed wth his re-usable system?

What is NELS?
[/quote]

http://www.rocketeers.co.uk/node/2277

[adrianwyard]

I think it's fair to say the business case for Skylon is dependent on its existence enabling a greatly expanded launch services market. For a reusable vehicle like Skylon, high flight-rates are needed because they lead quickly to lower costs per kg to orbit, and therefore lower prices. Lower prices lead new customers in this expanding market to chose Skylon over other providers, and all this works in a nice positive feedback loop. At least that's the theory.

If all goes well, Skylon should offer a very respectable payload mass and volume to orbit. ~15 tonnes and 4.8m will seem modest when compared to SLS or Falcon Heavy, but it's comparable to many of todays launchers and capable of lofting most of the payloads on the order books.

If we fast-forward to the mid 2020s, Skylon supporters can hope that it's successfully captured a lot of the medium sized satellite launch market, but by then SLS and Falcon Heavy would presumably be available. Could it also launch some of the megaprojects they are slated to loft? There are a few cases where a 5m, 7m or, 10m  diameter fairing is a fundamental payload requirement, but what if this is not a driving factor, and the task is simply to launch a great deal of mass?

Something like the ISS is an interesting example. By mass alone, some 30 Skylon flights would be needed to rebuild the entire ISS (or something analogous). An order for 30 flights is exactly the sort of thing a fledgling Skylon carrier needs to begin operations. Curiously, SLS would not be very efficient at launching ISS-scale modules as their size is a poor fit. By mass (450 t) it would only take 5 Block IA SLS launches, but the ISS has >20 sizable components, and my guess is it would actually take 7-11. (Two ISS pressurized modules can fit side-by-side in a 10m fairing, and the 7 launch estimate assumes you also stack a third module above these two.)

This is admittedly a very contrived example. No-one is going to build a carbon copy of the ISS in 2025. But it's meant to set up the following question: if a megaproject does not absolutely require large diameters, are these better launched on SLS, or piecemeal on standard-sized launchers? If they can be launched by Skylon, this is a point in favor of the expanded market that they need (and one fewer job for SLS to do). If, however, they are better launched with SLS, this does not bode well for Skylon, (and it's a job that SLS retains).

For megaprojects that bear a resemblance to the ISS, my guess is an order of 30 Skylon flights will cost less than 7 Block IA SLS.

Some related points:
+ Unlike the ISS project that used the Shuttle for assembly (US segment) both Skylon and SLS would require automated rendezvous and docking technology if the project was not a single unit.)
+ For a project with many launches, the cost of the automated docking technology would be highly diluted.
+ A project that assumes SLS can be launched with many parts integrated and ground-tested. (Skylab-style).
+ A project launched in smaller pieces endures lower launch loads.
+ A project launched in smaller pieces can recover more easily if a component is lost on the way to the assembly site, or if it fails on orbit.
+ A project launched in smaller pieces can be upgraded more easily.
+ I am ignoring the realities of international politics and trade - e.g. in reality a US Gov project is not likely to buy Skylon flights, and of course other counties cannot simply buy SLS flights. I'm just trying to think about the generalized arguments people make for small/cheap/frequent services vs super-heavy/integrated approaches to launching big projects.
 
 I realize there are too many unknowns here to really argue to conclusions, but I'd welcome any responses.

[Jim Davis]

I realize there are too many unknowns here to really argue to conclusions, but I'd welcome any responses.

If, for the sake of argument, the 10 billion in dollars, euros, pounds sterling or whatever currency becomes available for an advanced launch system the chances that it is going to be Skylon are almost zero. When that kind of money becomes available it invariably goes to organizations that have experience in managing projects of that size, a group that does not include Reaction Engines. Inevitably, the European airframe (BAE Systems, EADS, etc) and engine (Rolls-Royce, SNECMA, etc) organizations, whether in competition or collaboration, are going to come up with a design or designs that will be completely unrelated to Skylon or SABRE. They may or may not use some technology developed by REL (heat exchangers). Small organizations like REL do not get 11 figure projects to manage. This is simple political reality.

Even less likely is that the money will be raised by private investment. Even projects with similar budgets and with far more mature enabling technology can be very risky, like the channel tunnel, for example.

I think (but would love to proven wrong) that 20 or 30 years from now Skylon will be viewed in the same manner we view von Braun's three stage ferry rocket, or Bono's aerospike SSTO RLVs, or Hudson's Roton -  interesting concepts, but well beyond the contemporary state of the art. 


 

[aero]

I'm wondering why it costs $10 billion to get something flying. A full scale proof of concept vehicle  perhaps. Skylon is afterall just a drone. A very large drone, but a drone none-the-less.

[mmeijeri]

I think it's fair to say the business case for Skylon is dependent on its existence enabling a greatly expanded launch services market. For a reusable vehicle like Skylon, high flight-rates are needed because they lead quickly to lower costs per kg to orbit, and therefore lower prices. Lower prices lead new customers in this expanding market to chose Skylon over other providers, and all this works in a nice positive feedback loop. At least that's the theory.

Skylon is so large because it aims to be profitable on the existing market. I'm skeptical about that, but who knows, Musk also believes he can do it that way. A large market seems more helpful to me, but only if it appears, which doesn't seem likely in the near future.

But RLVs aren't expected to be profitable until they can get flight rates of about 50 / yr (the numbers vary from pundit to pundit). Even for a very small RLV, you'd need hundreds of tonnes of payload a year, especially if you want competing suppliers. That's about the amount a serious exploration program would give you. And conveniently most of that mass would be cheap propellant, not expensive aerospace hardware, and most of that hardware could be reusable. All in all, this should be technically and financially possible, though probably not politically possible.

If you want to do the same for a Skylon-sized RLV you'd need to launch thousands of tonnes a year, which seems unrealistic.

So there you have it, some believe Skylon could be profitable on the existing market precisely because it is relatively large, others believe it is impossibly large and cannot hope to be profitable.

[anonymous]

If, for the sake of argument, the 10 billion in dollars, euros, pounds sterling or whatever currency becomes available for an advanced launch system the chances that it is going to be Skylon are almost zero. When that kind of money becomes available it invariably goes to organizations that have experience in managing projects of that size, a group that does not include Reaction Engines. Inevitably, the European airframe (BAE Systems, EADS, etc) and engine (Rolls-Royce, SNECMA, etc) organizations, whether in competition or collaboration, are going to come up with a design or designs that will be completely unrelated to Skylon or SABRE. They may or may not use some technology developed by REL (heat exchangers). Small organizations like REL do not get 11 figure projects to manage. This is simple political reality.

REL isn't asking to run the project or manufacture Skylon themselves. They say that would be done by the major aerospace companies. All REL would manufacture would be the pre-cooler. REL is designing Skylon now in order to show that the idea could work, not because it would be the final design, but it would be stupid for the major companies to throw all that work away and start designing from scratch. If you're going to criticise Skylon, get your facts right. Don't put up straw man arguments that only show you can't be bothered.