The Reaction Engines Skylon Master Thread (5)

[Chris Bergin]

Fifth thread for Reaction Engines/Skylon.

Previous: The Reaction Engines Skylon Master Thread (4)
http://forum.nasaspaceflight.com/index.php?topic=34964.0

This has to be on topic and civil. This is for sensible debate and updates. Anything trivial or stupid will be deleted without notice.

[jrc14]

Just to clarify a point on 'yield' that seems to be causing misunderstanding, and responding to recent posts (in thread #4) by t43562, francesco nicoli and others ...

Why I think I know the answer: I have been working in investment banks building quant models (including models of bonds and other investments) for the past 25 years
Why it's important: thanks, in part, to imprecise wording in the investopedia article on 'Yield', forum readers might misunderstand how 'yield' is calculated - and so they'll get a wildly misleading view of how the Skylon business case works (or does not work)
What the answer is: when people quote 'bond yield' or 'return on investment' or 'interest' as a percentage figure, that is always an annual figure (its units being "quantity per cent per year").
The answer in tedious detail:
Suppose that somebody issues, and that I buy, a ten year bond paying a yield of 8%, and that I buy $100 of this bond.  The financial outcome is:

I pay today $100
I receive 1 year from now $8 (interest, also known as coupon)
I receive 2 years from now $8 (interest, also known as coupon)
...
I receive 10 years from now $108 (being repayment of my initial $100 plus $8, being the final interest payment)

So overall, I have paid out $100 and I have received $180.  Deciding whether this was a fair deal (for me or for the bond issuer) is a complex question, and for those wanting an answer the Wikipedia article on Discounting is as good a place to start as any.
Bank account interest, loans, mortgages and so forth all follow this general pattern.  The interest rate is an annual amount, describing how much interest is paid each year.

[Note: to the pedants out there: yes, I am aware of semi-annual interest, compounding, daycount convention, the definition of IRR, the difference between nominal/coupon yield, current/OTR yield and redemption yield, the roles of exchanges, depositories and clearing houses, ... . I have simplified the explanation, for the sake of exposition.  If you know enough to challenge me on any of these details, you also know enough to agree with my central point, which is that yield is an annual amount]

[Impaler]

jrc14:  Thank you for for that well written post.  I hope this settle the matter for anyone still in doubt and we may now discuss how Skylon business case closes/dose-not-close using correct financial principles.

jrc14, can you give us some rough estimates of what a potential Skylon operators cost of capitol might be and thus generate some estimates of what kind of gross profit is needed over a range of time-frames to breakeven.

Lets assume 1 Billion is the initial outlay for the Skylon, and time periods are 2 year, 5, years, 10 years, 20 years.  How much do they need to make over and above operating costs per year and over the period in question to justify the investment?

[t43562]

Just to clarify a point on 'yield' that seems to be causing misunderstanding, and responding to recent posts (in thread #4) by t43562, francesco nicoli and others ...
What the answer is: when people quote 'bond yield' or 'return on investment' or 'interest' as a percentage figure, that is always an annual figure (its units being "quantity per cent per year").

Thanks for the explanation.

[ChrisWilson68]

To start off the new thread, here's a summary of my own reasons for being skeptical about Skylon.  Some of these views are probably shared by other skeptics.  Feel free to reply with opposing views; hopefully, this will help clarify exactly where opinions differ and help undecided readers of these forums see both sides and make up their own minds.

First off, I don't think there's a known flaw in Skylon that definitely makes it impossible.  It's not like a perpetual motion machine that violates known laws of physics.  My issue with Skylon is that there are too many unknowns and the proponents of Skylon assume those unknowns will work out, while history shows this is seldom the case.  There are enough unknowns and enough projections that seem very optimistic to me that the odds of Skylon actually achieving its goals seem remote to me.

The people working on Skylon have been working on the idea for decades.  They are certainly dedicated and well-meaning, and they have some competence.  But they have been working on theory and small components.  They don't have experience in system integration.  They haven't built real flight hardware.  They haven't seen a system from concept through to all the inevitable compromises necessary to make a practical system.

The team at REL has consistently proposed building a large-scale single-stage-to-orbit system.  That shows poor judgement, in my opinion.  SpaceX started with Falcon 1.  Then then moved to a full-expendable Falcon 9.  Now they are working on perfecting reuse of only the first stage.  Along the way, they have learned many lessons and constantly changed their plans, all while retaining their goal of greatly reducing the cost of launch.  I believe that kind of incremental, flexible approach is very effective.  It is the opposite of the REL approach.  With REL going directly for a huge, single-stage-to-orbit system, there is little room to learn operational lessons and change plans.  And Skylon is so much different from existing systems it is very likely to need far more flexibility for lessons learned than Falcon.

There have been many programs with similar or lesser optimistic goals that have failed.  The U.S. National Aerospace Plane had far more resources available and a similar level of technological challenge, and it failed.  Note that I'm not saying the details of the technological challenge are similar -- they are not.  But the programs are similar in having a goal that required many unknowns to be overcome and having people with some competence in specific areas convinced they could overcome them.

The X-33/VentureStar is another launch program that had optimistic goals and failed.  I think that X-33/VentureStar looked far more realistic at its outset, with less of a techonological leap required, than Skylon today.  And yet it failed because of the engineering details in turning the theory into reality.

I also find the projected business model of Skylon implausible.  It posits 30 units of Skylon will be bought for a billion dollars each.  That would give a launch capability far, far beyond the current market, at a price not enough lower to justify the enormous market expansion.  One commonly-used cost figure is $5 million per flight based on 200 flights per vehicle and 30 vehicles.  If the market were really there for such a launch rate, SpaceX could develop a fully-reusable upper stage for Falcon 9 and cover it at an even lower cost.

Another part of the business case is that governments will by Skylon units for prestige.  I think that's unrealistic because national space programs get prestige from developing indigenous capabilities far more than by buying from another country.

So, there you have it.  If you disagree, post what you disagree with and why.  If you're a reader and undecided, read this and the responses and make up your own mind.

One final note: I hope I'm wrong and that Skylon beats the odds and succeeds.  But hope shouldn't mean we aren't realistic about how unlikely something is.

[t43562]

The people working on Skylon have been working on the idea for decades.  They are certainly dedicated and well-meaning, and they have some competence.  But they have been working on theory and small components.  They don't have experience in system integration.  They haven't built real flight hardware.  They haven't seen a system from concept through to all the inevitable compromises necessary to make a practical system.

I don't have a biography to hand of each one of them but I suggest that it's a sweeping statement to say that they have no experience in system integration.  At the very least some of them are veterans of Blue Streak and others of the aerospace industry.

The team at REL has consistently proposed building a large-scale single-stage-to-orbit system.  That shows poor judgement, in my opinion.  SpaceX started with Falcon 1.  Then then moved to a full-expendable Falcon 9.  Now they are working on perfecting reuse of only the first stage.  Along the way, they have learned many lessons and constantly changed their plans, all while retaining their goal of greatly reducing the cost of launch.  I believe that kind of incremental, flexible approach is very effective.  It is the opposite of the REL approach.  With REL going directly for a huge, single-stage-to-orbit system, there is little room to learn operational lessons and change plans.  And Skylon is so much different from existing systems it is very likely to need far more flexibility for lessons learned than Falcon.

If my memory serves me their contention is that a smaller system does not offer the economic benefits. In other words the nature of the problem dictates the way they have to go - they can't scale up over time.  Assuming this is true, what do you expect them to do?

I think it's also fair to say that if they can build a demonstrator engine they'll be able to remake their decisions having learned from it. I don't see them claiming that their plan is set in stone. If you consider HOTOL, they have already learned and changed quite a lot.

[Lars-J]

The people working on Skylon have been working on the idea for decades.  They are certainly dedicated and well-meaning, and they have some competence.  But they have been working on theory and small components.  They don't have experience in system integration.  They haven't built real flight hardware.  They haven't seen a system from concept through to all the inevitable compromises necessary to make a practical system.

I don't have a biography to hand of each one of them but I suggest that it's a sweeping statement to say that they have no experience in system integration.  At the very least some of them are veterans of Blue Streak and others of the aerospace industry.

But that goes for almost any aerospace start-up. Among its employees you would expect to see some veterans from other aerospace firms - but that does not mean that the "organizational experience" as a whole translates to this new organization. 

[t43562]

The people working on Skylon have been working on the idea for decades.  They are certainly dedicated and well-meaning, and they have some competence.  But they have been working on theory and small components.  They don't have experience in system integration.  They haven't built real flight hardware.  They haven't seen a system from concept through to all the inevitable compromises necessary to make a practical system.

I don't have a biography to hand of each one of them but I suggest that it's a sweeping statement to say that they have no experience in system integration.  At the very least some of them are veterans of Blue Streak and others of the aerospace industry.

But that goes for almost any aerospace start-up. Among its employees you would expect to see some veterans from other aerospace firms - but that does not mean that the "organizational experience" as a whole translates to this new organization.

That is a different argument - if we are respecting the English language and the use of absolutes.

They are an engine company not an airframe company.   Let the airframe company use its great institutional experience for its part of the work.

[ChrisWilson68]

The people working on Skylon have been working on the idea for decades.  They are certainly dedicated and well-meaning, and they have some competence.  But they have been working on theory and small components.  They don't have experience in system integration.  They haven't built real flight hardware.  They haven't seen a system from concept through to all the inevitable compromises necessary to make a practical system.

I don't have a biography to hand of each one of them but I suggest that it's a sweeping statement to say that they have no experience in system integration.  At the very least some of them are veterans of Blue Streak and others of the aerospace industry.

But that goes for almost any aerospace start-up. Among its employees you would expect to see some veterans from other aerospace firms - but that does not mean that the "organizational experience" as a whole translates to this new organization.

That is a different argument - if we are respecting the English language and the use of absolutes.

They are an engine company not an airframe company.   Let the airframe company use its great institutional experience for its part of the work.

That's fine, but this engine company is making projections about the performance and economic viability of the complete system, including engines and airframe.

[t43562]

The people working on Skylon have been working on the idea for decades.  They are certainly dedicated and well-meaning, and they have some competence.  But they have been working on theory and small components.  They don't have experience in system integration.  They haven't built real flight hardware.  They haven't seen a system from concept through to all the inevitable compromises necessary to make a practical system.

I don't have a biography to hand of each one of them but I suggest that it's a sweeping statement to say that they have no experience in system integration.  At the very least some of them are veterans of Blue Streak and others of the aerospace industry.

But that goes for almost any aerospace start-up. Among its employees you would expect to see some veterans from other aerospace firms - but that does not mean that the "organizational experience" as a whole translates to this new organization.

That is a different argument - if we are respecting the English language and the use of absolutes.

They are an engine company not an airframe company.   Let the airframe company use its great institutional experience for its part of the work.

That's fine, but this engine company is making projections about the performance and economic viability of the complete system, including engines and airframe.

Don't they have to? Is it not necessary at all times to make such projections and keep updating them as new information is learned? Would anyone even bother to invest if the projections of today were bad? It doesn't mean their projections are 'plucked out of the air' or based on nothing.  So they will build their test engine and perhaps other things on the way and then their projections will have much more authority whatever they turn out to be but for now we have what we have and can only discuss that.

[ChrisWilson68]

That's fine, but this engine company is making projections about the performance and economic viability of the complete system, including engines and airframe.

Don't they have to? Is it not necessary at all times to make such projections and keep updating them as new information is learned?

There'd be no shame in their saying "we don't know yet".

Would anyone even bother to invest if the projections of today were bad?

Lots of people make lots of bad investments.  The mere fact that someone has invested in not sufficient to overrule our judgement if we have reason to doubt a projection.

The projections of the early Space Shuttle studies were way off.  The projections of the X-33/VentureStar project were way off.  And in both of those cases you had organizations with much more relevant experience.

[lkm]

To start off the new thread, here's a summary of my own reasons for being skeptical about Skylon.  Some of these views are probably shared by other skeptics.  Feel free to reply with opposing views; hopefully, this will help clarify exactly where opinions differ and help undecided readers of these forums see both sides and make up their own minds.

First off, I don't think there's a known flaw in Skylon that definitely makes it impossible.  It's not like a perpetual motion machine that violates known laws of physics.  My issue with Skylon is that there are too many unknowns and the proponents of Skylon assume those unknowns will work out, while history shows this is seldom the case.  There are enough unknowns and enough projections that seem very optimistic to me that the odds of Skylon actually achieving its goals seem remote to me.

The people working on Skylon have been working on the idea for decades.  They are certainly dedicated and well-meaning, and they have some competence.  But they have been working on theory and small components.  They don't have experience in system integration.  They haven't built real flight hardware.  They haven't seen a system from concept through to all the inevitable compromises necessary to make a practical system.

I don't think you are fully aware of the background the REL team. Mark Hempsell for example worked on the DCX.
http://forum.nasaspaceflight.com/index.php?topic=19127.msg493088#msg493088

The team at REL has consistently proposed building a large-scale single-stage-to-orbit system.  That shows poor judgement, in my opinion.  SpaceX started with Falcon 1.  Then then moved to a full-expendable Falcon 9.  Now they are working on perfecting reuse of only the first stage.  Along the way, they have learned many lessons and constantly changed their plans, all while retaining their goal of greatly reducing the cost of launch.  I believe that kind of incremental, flexible approach is very effective.  It is the opposite of the REL approach.  With REL going directly for a huge, single-stage-to-orbit system, there is little room to learn operational lessons and change plans.  And Skylon is so much different from existing systems it is very likely to need far more flexibility for lessons learned than Falcon.

SABRE is a SSTO engine, I'm not sure what intermediate stage there can be for engine explicitly designed to take a single stage into orbit. I can't imagine designs cost get significantly smaller by making a smaller version.

There have been many programs with similar or lesser optimistic goals that have failed.  The U.S. National Aerospace Plane had far more resources available and a similar level of technological challenge, and it failed.  Note that I'm not saying the details of the technological challenge are similar -- they are not.  But the programs are similar in having a goal that required many unknowns to be overcome and having people with some competence in specific areas convinced they could overcome them.

The technical challenge of airbreathing to Mach 18 is clearly of a vastly higher level than airbreathing to Mach 5.5 and the number of unknowns in geting to Mach 18 in 1984 far greater than achieving Mach 5 thirty years later.

The X-33/VentureStar is another launch program that had optimistic goals and failed.  I think that X-33/VentureStar looked far more realistic at its outset, with less of a techonological leap required, than Skylon today.  And yet it failed because of the engineering details in turning the theory into reality.

X-33 didn't fail, it was cancelled due to a change in administration, just as many Clinton era space programs were cancelled by the Bush administration. What many people forget is that the X-33 was just a rocket powered x-plane like the X-15, like the X-15 it had a ton of not flown before technology some of which had teething problems, and like the X-55 if it had flown it would have provided invaluable hypersonic flight data.

I also find the projected business model of Skylon implausible.  It posits 30 units of Skylon will be bought for a billion dollars each.  That would give a launch capability far, far beyond the current market, at a price not enough lower to justify the enormous market expansion.  One commonly-used cost figure is $5 million per flight based on 200 flights per vehicle and 30 vehicles.  If the market were really there for such a launch rate, SpaceX could develop a fully-reusable upper stage for Falcon 9 and cover it at an even lower cost.

Your comparing apples to oranges, the current launch capability is already far, far, larger than the current launch market and yet it is sustained by it. Further 30 Skylons are not just going to come into existence upon the commencement of commercial availability, they have to be built. At a rate of 2 a year there wouldn't be 30 Skylon's arround until 2040, 35 years from the current market. Finally the question that needs to be asked isn't how many launches could be made each year but how few could be made while still turning a profit at a competitive price because if a profit can be made nothing else matters.What seems to be the case is that at an incremental launch cost of $5 million a very high launch rate can spread fixed costs so widely that the market launch price can be not much more than that or at a much lower launch rate the market price can be competitive with current launchers.

Another part of the business case is that governments will by Skylon units for prestige.  I think that's unrealistic because national space programs get prestige from developing indigenous capabilities far more than by buying from another country.

The business case for governments is massively more responsive space access and, for currently non-space-faring countries, a greatly increased asssured access to space and sovereign space capability as well the economic development that comes with having such assets akin to governments having national flag airlines, building massive hub airports and ports etc.
 

So, there you have it.  If you disagree, post what you disagree with and why.  If you're a reader and undecided, read this and the responses and make up your own mind.

One final note: I hope I'm wrong and that Skylon beats the odds and succeeds.  But hope shouldn't mean we aren't realistic about how unlikely something is.

Personally I think the greatest threat to Skylon development will be a failure of an airframer to commit to it. I think REL may find that despite developing an engine that works well and engenders a lot of interest in the end there may be a general reticence to throw in with another companies grand scheme and disrupt their own planning among the likely consortium partners. I could see REL being bought by RR as a part of an attempt at forming a successful consortium only to end being used to get some lucrative US hypersonics research money.

[ChrisWilson68]

To start off the new thread, here's a summary of my own reasons for being skeptical about Skylon.  Some of these views are probably shared by other skeptics.  Feel free to reply with opposing views; hopefully, this will help clarify exactly where opinions differ and help undecided readers of these forums see both sides and make up their own minds.

First off, I don't think there's a known flaw in Skylon that definitely makes it impossible.  It's not like a perpetual motion machine that violates known laws of physics.  My issue with Skylon is that there are too many unknowns and the proponents of Skylon assume those unknowns will work out, while history shows this is seldom the case.  There are enough unknowns and enough projections that seem very optimistic to me that the odds of Skylon actually achieving its goals seem remote to me.

The people working on Skylon have been working on the idea for decades.  They are certainly dedicated and well-meaning, and they have some competence.  But they have been working on theory and small components.  They don't have experience in system integration.  They haven't built real flight hardware.  They haven't seen a system from concept through to all the inevitable compromises necessary to make a practical system.

I don't think you are fully aware of the background the REL team. Mark Hempsell for example worked on the DCX.
http://forum.nasaspaceflight.com/index.php?topic=19127.msg493088#msg493088

Actually, that demonstrates my point quite well.  DC-X never progressed to an operational vehicle.  It was never more than a sub-scale technology demonstrator.  Working on DC-X wouldn't give any experience with the very difficult transition from technology demonstration to a system that is economically successful as an operational system.

And "worked on" isn't the same as being in charge.

The team at REL has consistently proposed building a large-scale single-stage-to-orbit system.  That shows poor judgement, in my opinion.  SpaceX started with Falcon 1.  Then then moved to a full-expendable Falcon 9.  Now they are working on perfecting reuse of only the first stage.  Along the way, they have learned many lessons and constantly changed their plans, all while retaining their goal of greatly reducing the cost of launch.  I believe that kind of incremental, flexible approach is very effective.  It is the opposite of the REL approach.  With REL going directly for a huge, single-stage-to-orbit system, there is little room to learn operational lessons and change plans.  And Skylon is so much different from existing systems it is very likely to need far more flexibility for lessons learned than Falcon.

SABRE is a SSTO engine, I'm not sure what intermediate stage there can be for engine explicitly designed to take a single stage into orbit. I can't imagine designs cost get significantly smaller by making a smaller version.

REL is the one who chose to design it for an SSTO vehicle.  They chose to optimize it for that role rather than as part of a reusable first stage of a two-stage launch system, which would be the more conservative choice and give them more margin and require much less in the way of pushing the edge of what technology can do.

And everything gets cheaper when things are smaller.

There have been many programs with similar or lesser optimistic goals that have failed.  The U.S. National Aerospace Plane had far more resources available and a similar level of technological challenge, and it failed.  Note that I'm not saying the details of the technological challenge are similar -- they are not.  But the programs are similar in having a goal that required many unknowns to be overcome and having people with some competence in specific areas convinced they could overcome them.

The technical challenge of airbreathing to Mach 18 is clearly of a vastly higher level than airbreathing to Mach 5.5 and the number of unknowns in geting to Mach 18 in 1984 far greater than achieving Mach 5 thirty years later.

Getting to Mach 5.5 isn't the challenge.  They still have to get to Mach 25 to make orbit.  True, in some ways it's easier if they're only air breathing to Mach 5.5.  But in other ways it's harder.  They have to carry much more oxidizer, and their engine has to work well in both air-breathing and rocket mode.  Going from Mach 5.5 to Mach 25 in rocket mode (with some of that rocket mode in the dense part of the atmosphere at Mach 5.5) means they need a very good mass fraction.

Like I said, they don't have exactly the same challenges NASP had, but they have very great challenges, and I think they're at a comparable level of difficulty.  Apparently, others think they are too, which is why others are continuing SCRAMJet research and development.

The X-33/VentureStar is another launch program that had optimistic goals and failed.  I think that X-33/VentureStar looked far more realistic at its outset, with less of a techonological leap required, than Skylon today.  And yet it failed because of the engineering details in turning the theory into reality.

X-33 didn't fail, it was cancelled due to a change in administration, just as many Clinton era space programs were cancelled by the Bush administration. What many people forget is that the X-33 was just a rocket powered x-plane like the X-15, like the X-15 it had a ton of not flown before technology some of which had teething problems, and like the X-55 if it had flown it would have provided invaluable hypersonic flight data.

Getting into the details of X-33 is off topic, but lets just say that opinions differ about X-33 -- many people believe it was cancelled because the progress up to that point indicated it was not able to meet its projections and wouldn't have been viable.

[SleeperService]

Are you one of those people that claim it's only 5% of the delta v to orbit?

[aga]

economic viability of the complete system, including engines and airframe.

that was done by others, not by rel... eg. esa, london economics, etc

[ChrisWilson68]

economic viability of the complete system, including engines and airframe.

that was done by others, not by rel... eg. esa, london economics, etc

It was done by REL.  All the projections come from REL.  A small group from ESA did a short audit of the REL plans to see if there was an obvious show-stopper.  They said there wasn't a showstopper they could see.  They didn't claim to know REL's projections were all correct.

[john smith 19]

REL is the one who chose to design it for an SSTO vehicle.  They chose to optimize it for that role rather than as part of a reusable first stage of a two-stage launch system, which would be the more conservative choice and give them more margin and require much less in the way of pushing the edge of what technology can do.

Which would in fact have doubled the development budget as they would had to look at 2 vehicles and the testing around hypersonic separation. The last attempt at which was IIRC the SR71/M4 drone separation tests.

This is from the man who's worried about "too many unknowns." 

You need to stop making assertions as fact. They make you look untrustworthy.

And everything gets cheaper when things are smaller.

You work in a Silicon valley start up and you aren't aware of price inflation between the last and next generation of wafer fabs as they've gone from about about 33 to 14nm?

Or the time and effort involved when you go from the piping in the 1960's Aerospace Plane (about 1cm) to the 1mm used in REL HX's.

Or REL estimate that a scaled down LH2 turbo pump (to give the same chamber pressure) for a small scale SABRE would eat abut £250m of the budget alone as engineering bearings to run about 12x faster than the full scale unit (in LH2) is much harder.

No smaller often means cheaper, but not always. Smart engineers are aware of this and plan accordingly.

[Paul451]

I don't see them claiming that their plan is set in stone.

Claiming, no. Acting, yes. They are designing payloads and space stations and Mars missions around the size/capacity of the Skylon payload bay. (Not to mention a hypersonic passenger plane.) And Bond rejects out of hand any suggestions that Skylon may not be the optimal design (as John echoes, above). If you were an "airframer", would you get mixed up with an engine company which behaves like that? Or wait until they fail and just licence the engines from whoever buys the IP, and develop your own clean-sheet design?

To me, it's like some who wants to develop the world's first jet engine. So far they have one compressor fan. But they've not only designed the rest of the engine, and designed the entire airliner around that engine, and insisted it's the only possible design, but they are proposing new airport designs based around the door spacing on that proposed airliner for the proposed engine for which they have (after 20 years) only built a single compressor fan.

But you dare suggest they are being a bit premature...

REL is the one who chose to design it for an SSTO vehicle. They chose to optimize it for that role rather than as part of a reusable first stage of a two-stage launch system

Which would in fact have doubled the development budget as they would had to look at 2 vehicles

{sigh} Why is this idea so prevalent in aerospace?

Two is more than one, so therefore it must cost twice as much to develop an aircraft to carry freight between cities and a truck to ferry between individual customers and the airfreight terminals than to develop a single vehicle which can fly between cities but land directly on the customers' driveways. Must. Because two is more than one.

and the testing around hypersonic separation.

Only if they were stupid.

When Chris is suggesting smaller stepping stones, when that's the entire premise of his argument, why would you assume he would be suggesting the hardest possible version of TSTO?

[t43562]

I don't see them claiming that their plan is set in stone.

Claiming, no. Acting, yes. They are designing payloads and space stations and Mars missions around the size/capacity of the Skylon payload bay. (Not to mention a hypersonic passenger plane.) And Bond rejects out of hand any suggestions that Skylon may not be the optimal design (as John echoes, above). If you were an "airframer", would you get mixed up with an engine company which behaves like that? Or wait until they fail and just licence the engines from whoever buys the IP, and develop your own clean-sheet design?

If one looks with a hostile attitude one can twist quite normal behaviour into something sinister.

Engines do really determine what kind of aeroplane can be built - it has always been that way around in the aircraft world I believe (I'm no expert but reading the history of flight makes it pretty obvious). So yes, the airframer is going to have to accept that they can't make an airframe any size or any shape they like.  They still have freedom to ignore REL's design if they have to a reason to.  Why would they be sulky about something that has been a fact of life in the aerospace industry since the beginning?

REL designed Mars missions to show themselves that it was possible to do such a thing with the design.  If it hadn't been then that would have meant there was something basically wrong with the whole idea and it would have to be modified.

Similarly they have studied how to launch satellites and boost them into GEO. Without doing this, who would invest? It's part of designing something that you put it through it's imaginary paces before you build it rather than afterwards. Surely? Would you consider it responsible not to? Surely this has to be done repeatedly with more and more detail until the final article does it for real?

[Krevsin]

REL is the one who chose to design it for an SSTO vehicle.  They chose to optimize it for that role rather than as part of a reusable first stage of a two-stage launch system, which would be the more conservative choice and give them more margin and require much less in the way of pushing the edge of what technology can do.

Which would in fact have doubled the development budget as they would had to look at 2 vehicles and the testing around hypersonic separation. The last attempt at which was IIRC the SR71/M4 drone separation tests.

I'm not an expert on any of this so feel free to correct me, but wouldn't it make more sense to develop a TSTO Skylon in such a way that the skylon makes a suborbital hop and, once outside the brunt of the atmosphere, open the cargo bay and deploy an upper stage to which the payload is attached?

In fact, hasn't something similar to this concept been proposed in this study by Mark Hempsell?

It'd certainly be less complicated than staging inside the atmosphere at hypersonic velocities IMO.