The Reaction Engines Skylon Master Thread (1)

[john smith 19]

REL has demonstrated a strong willingness to plan ahead and work to international standards with some longevity.

One small point which is is obviously a long way out concerns the propellant loading.

This is planned to be automated, with the couplings in one of the wheel wells, eliminating the need for yet another set of doors/latches/actuators on the vehicle.

An interesting option would be to allow this wheel to drop on orbit, allowing access to these couplings. This would *potentially* allow on orbit refueling to take place.

Obviously this would never be able to land on anywhere else in the solar system but as a way of building an Earth departure stage for *no* development budget it would be an interesting possibility. Skylon then becomes the bus for a payload, which would probably have to do its own braking burn.

This is one of those applications that should be fairly simple to engineer in at the design stage but a nightmare to do so later on.

One question would be if it *was* possible to do a return trip from say the Moon would the tires be usable. STS tires were very conservatively designed. Modern designs should do better than 14 days on orbit with a single use. A key issue would be micrometeroid damage during refueling, when the wheels are not protected.

[simonbp]

Not sure if you answered this: Is there an altitude limit? I'd guess takeoff from Denver would not be feasible.

That is an interesting point. Maybe need a longer runway, but once the vehicle is airborne it shouldn't matter. Plus, not having to blast through as much lower atmosphere could actually increase performance.

[john smith 19]

Patchouli “I wonder will Skylon use something similar to what ULA is proposing for their Integrated Vehicle Fluids concept to solve the problem of on orbit power and supplying GOX and H2 to the RCS system?  One thing that seems would be fitting is use of a hydrogen oxygen fueled ICE for the APU.  Of course with Skylon being reusable the high cost of fuel cells is probably less of an issue then it is with an expendable upper stage.

Basically the answer is yes.  There is a gaseous feed created from the orbital propellant tanks which sends hydrogen and oxygen to the RCS thrusters, fuel cells and APU (although the APU might be replaced by bigger fuels cells and a more electric design approach to actuation).

This seems a little odd to me. I think the technology risk of using electric actuators has fallen a lot (driven by large aircraft but I'd include the
1st Vega launch this February as the first *big* rocket engine to rely on them for TVC).

 I'm less clear why the preference for fuel cells over say a micro-turbine based emergency power unit adapted to GH2/GO2. This adaptation would seem to play better to REL's strengths in combustion engineering.

I recall the gearing between the turbine and the pump on the Shuttle's APU's was a major maintenance issue, but these units use direct coupled generators and power electronics to give the right supply voltages and current in a more compact structure. The foil air bearings have a long history in aircraft "Air Cycle Machines" and have very low maintenance requirements. I believe these bearings have had a reputation for being a "Black art" but there is now substantial public domain information available on them.

When NASA looked at electric TVC for the SSME's they were looking at 57Kw actuators per axis per engine, a very big jump from the total of 36Kw from the 3 standard fuel cells (even in their 45Kw burst mode), even assisted by regenerative actuators and some kind of load leveling package.

I'm aware there have been EU fuel cell projects but I'm unaware of any for airborne or space based applications.

It did occur to me that SKylon might be very sensitive to turning torques generated by the spinning masses involved but I'd presume this could be counted by mounting 2 units back to back, provided they stayed synchronised . A more creative solution would be a dual contra-rotating APU

On costs I was hoping it would possible to give some idea of how much of that c£10Bn cost would be needed to separate out the cost of setting up the mfg plant and first flight vehicle from the cost of the test programme.

[flymetothemoon]

On costs I was hoping it would possible to give some idea of how much of that c£10Bn cost would be needed to separate out the cost of setting up the mfg plant and first flight vehicle from the cost of the test programme.

Me too. Even as a percentage of the total currently undefined updated cost. When asked about this in the recent interview Mr Hempsell said, "It depends" - on the cost of money at dev time as much as anything.

I know that the first airframes might not be orbital, so perhaps two figures would be interesting. The approximate (percentage) spend required for the first full-size test airframe and the approximate (percentage) spend required for the first full-size test orbital airframe.

As REL have already stated, operators might choose to start using the ships before the end of the complete aircraft qualification programme, so it would be interesting to know the earliest an operator could choose to do that "at their own risk".

[RobLynn]

Could Skylon ultimately be cheaper and faster for long distance travel?

These calculation are a little bit rough, but it looks like a low maintenance Skylon could ultimately do very well for suborbital antipodeal passenger flights, perhaps several time cheaper than the hypersonic Lapcat.

Lapcat uses about 200 tonnes of LH2 to fly 300 passengers to the antipodes (half way around the world).  Guessing 150 kg per passenger including seat and life support that is about 45 tonnes of payload, and about 666kg of hydrogen per passenger (costing several thousand dollars)

Skylon C2 uses about 63 tonnes of hydrogen to put 12 tonnes in LEO.  But a suborbital antipodeal trajectory supposedly saves about 1500m/s (perhaps more given high hypersonic L/D) adding about 23 tonnes of extra payload - call it 35 tonnes payload all up.  Scaling up slightly to 45 tonnes payload to match Lapcat with a bigger craft would need about 77 tonnes of hydrogen, (LOX costs are <10% of LH2 costs). or about 260kg LH2 per person.

Believe it or not that is very close to the same energy use per km over an antipodeal trip as for a commercial jet.

So if you could make a bullet-proof reliable Skylon with low maintenance costs then the fuel cost might only be about 40% of the Mach 5 Lapcat, and given flight times of about 1 hour your wage and capital costs would likely be even lower than 4 hour flight time Lapcat.

I believe re-entry heating is not much worse than for Mach 5 cruise, and the Sabre engines would spend far less time operating.  You could likely forego toilets, food etc and pack people in like sardines into a removable passenger module (though lots of sick bags required).

The short time in the stratosphere creating NOx and relatively short time creating supersonic booms would also make life a lot easier in terms of not annoying environmentalists.

1 Skylon could perhaps do the work of 6-10 conventional long range jets of the same capacity.  And operating costs might eventually be not much higher than for conventional subsonic Jets providing maintenance cost are low.

Interestingly replacing the 2 Sabres with 4x SSME type engines in a suborbital Skylon type vehicle could potentially do about the same if the engines were sufficiently cheap to maintain.  Assuming 4200m/s Isp and 7500m/s deltaV gives a mass ratio of 6, so the 44 tonne empty Skylon + 45 tonne passenger payload would have a GTOW of 540 tonnes and would need tanks very slightly bigger than Skylon C2 with the same 75 tonnes of LH2 fuel (250kg per passenger).  Structure would need to be a little heavier, but 4xSSMEs are about 6 tonnes lighter than 2x Sabres, so probably not much in it.

[ARD]

What about noise?  The SABRE engines combined produce about 2/3 the thrust of a Delta IV--if noise is a function of thrust, then this will be considerably louder than any other airliner.  Just a little bit upthread, Skylon's noise level was described as "Space Shuttle level."  Unless SABRE can deep-throttle (assuming that affects noise levels greatly), Skylon won't be operating from major commercial airports.  LAPCAT A2, OTOH, is designed to fly subsonic at times--presumably, no louder than the A380 is supposed to be. 

On the other hand, using Skylon as a sub-orbital airliner would do wonders for the flight rate. 

[john smith 19]

Believe it or not that is very close to the same energy use per km over an antipodeal trip as for a commercial jet.

It's something people often don't seem aware of. It was first noticed about 1/2 a century ago by Philip Bone at Douglas and was one of the inspirations for his series of "Sub Orbital Transports" However he thought that only VTOL and plug nozzle engines would make it viable.

Hempsell has mentioned that a Skylon on suborbital trajectory would carry about 30 tonnes although he seemed doubtful there would be much use for one. BTW last time I checked LH2 was about $5/lb and LO2 $0.15/lb.

However the takeoff noise may well be a killer. Takeoff noise for the Shuttle was c 170dB (well above the human pain threshold). Most of it was provided by the SRBs. Modern turbofans are *much* quieter (and larger) than the old turbojets because they mix a *lot* of low speed air in from the outside with the small fast moving core.

*If* SABRE does the same at ground level then the noise level will be more the difference between the *bulk* airflow through the engine and the surrounding atmosphere will set the noise level. If SABRE accelerates that flow to near *core* velocity then it's likely to get loud.

There is also the little matter that core gas velocity is that of a LH2/LO2 rocket, *not* a turbine. I'm not sure what that is on a gas turbine, but I think it's *substantially* lower than the 4-5000 m/s^-1 of a rocket, and AFAIK that is a *tough* parameter to change as it's a function of combustion chamber geometry.

However following REL's policy that once you buy one it's your choice what you do with it they *could* include a flight mode in the computer software to allow it. The thing is while the proverbial eccentric billionaire could have a place big enough that the noise would not upset the neighbors on take off, where would they *go*? Even substantially below full orbital payload it still takes a *very* serious runway to operate from.

[RobLynn]

Hempsell has mentioned that a Skylon on suborbital trajectory would carry about 30 tonnes although he seemed doubtful there would be much use for one. BTW last time I checked LH2 was about $5/lb and LO2 $0.15/lb.

However the takeoff noise may well be a killer. Takeoff noise for the Shuttle was c 170dB (well above the human pain threshold). Most of it was provided by the SRBs. Modern turbofans are *much* quieter (and larger) than the old turbojets because they mix a *lot* of low speed air in from the outside with the small fast moving core.

*If* SABRE does the same at ground level then the noise level will be more the difference between the *bulk* airflow through the engine and the surrounding atmosphere will set the noise level. If SABRE accelerates that flow to near *core* velocity then it's likely to get loud.

I agree noise is a huge problem, not many ways around it, though a catapult launch could help a lot, and with the fuel saved by a catapult you might be able to run the engines at about 20-30% of full thrust for a couple of minutes (about 10km per minute) until well out over sea before ramping up to full power.

Rock bottom natural gas prices as a result of shale gas mean LH2 prices for a big user can probably be cheaper than kerosene these days.  Natural gas can be converted to hydrogen with about 80% efficiency and currently costs about $3 per GJ.  Making LH2 including liquefaction takes about 200-250MJ worth of Natural gas per kg of LH2 so that is about $0.75 per kg in feedstock costs.  Kerosene is close to $1/kg.

So with a large enough demand and some competition in the supply chain LH2 could be very cheap - again making hypersonic or suborbital transport much more economic.

[adrianwyard]

Using Skylon for point-to-point was raised a few pages back. Mark Hempsell replied as follows:

The answer is it (sub-orbital point-to-point) isn’t easier – nor is the business case (for high cost, high-speed civil transport) yet clear.

(Parenthetical inserts are mine).

I was initially surprised at this, but when you start to think about launch-site requirements (noise, special runway - perhaps just one or two in the world - fuel infrastructure) then the space launch market does start to look more realistic.

[adrianwyard]

Continuing to worry a little about rocket plumes near runways and the rear fuselage: Even if the engines were gimballed to point horizontally during the runway acceleration, at rotation they'd be pointing toward it, or even gimballed downwards to gain alititude quickly. However, while near the runway the plume is going to reflect back off of the runway. Obviously, this reflected plume needs to be kept well clear of the rear Skylon structure.  Perhaps they'd be gimballed outward, away from the rear fuselage until at altitude... But even then that plume is going to expand as you ascend into thinner air. I imagine this sort of problem is fairly easy to model, so hopefully our friends at REL have it in hand.

I've wondered if the takeoff will look a bit more rocket-like than how it's depicted in the animations. In those the exhaust plume looks rather like a Concorde take-off where the reheat just added those nice Mach diamonds. However, with Skylon we need to picture eight rocket engine bells with an exhaust similar to that of the Shuttle SSMEs... Quite impressive! And it's probably a good thing the runway is re-inforced as these engines are pointing at it...

Now I think about it the rockets could presumably be gimballed to point directly aft until V1 speed, and only pointed Earthward at the last minute to get airborne, and at that speed the plume doesn't hit much runway area for very long.

And getting back to noise: my suspicion is that anyone who buys a ticket for a Skylon passenger module will be handed some nice heavy duty noise-canceling headphones for the takeoff.

[93143]

There is also the little matter that core gas velocity is that of a LH2/LO2 rocket, *not* a turbine.

Well...  in airbreathing mode it's a hydrogen/air rocket, not a hydrogen/oxygen rocket.  So I would expect that to be a bit lower.

[Robotbeat]

There is also the little matter that core gas velocity is that of a LH2/LO2 rocket, *not* a turbine.

Well...  in airbreathing mode it's a hydrogen/air rocket, not a hydrogen/oxygen rocket.  So I would expect that to be a bit lower.

It's still supersonic.

[93143]

Of course.  How do you suppose they'd get past Mach 5 with subsonic exhaust?

Using CEA, with a chamber pressure of 2000 psia, an oxynitrogen surrogate atmosphere (okay, I was lazy) at 123 K, and a 1:6 mix of 200 K hydrogen to oxygen by mass, the exhaust velocity appears to be in the vicinity of 2500 m/s at altitude, or 2240 for ideal expansion at sea level.  Probably around Mach 6 in the ambient air over the runway, even for the real system...

SSME seems to be in excess of Mach 10 ambient at sea level...

[john smith 19]

Of course.  How do you suppose they'd get past Mach 5 with subsonic exhaust?

Using CEA, with a chamber pressure of 2000 psia, an oxynitrogen surrogate atmosphere (okay, I was lazy) at 123 K, and a 1:6 mix of 200 K hydrogen to oxygen by mass, the exhaust velocity appears to be in the vicinity of 2500 m/s at altitude, or 2240 for ideal expansion at sea level.  Probably around Mach 6 in the ambient air over the runway, even for the real system...

SSME seems to be in excess of Mach 10 ambient at sea level...

Interesting points. I'd thought SABRE in air breathing mode was running at around 100 atm c1500psia. The point it's nearer M6 rather than M10 should also help as IIRC the correlation of exhaust velocity to actual noise level is *very* non linear. I *dimly* recall something V^4 but it's been a long time since I checked it.

Like turbofans I ground noise will depend how much of the airflow is fed to the combustion chambers and how much gets by passed around them.

Incidentally did that run give any idea what the plume size would be? People has expressed concern over plum impact on Skylons tail. I think this is highly unlikely given the engine locations on Skylon *unless* the plume is huge.

So far it looks fair to say that a Skylon take off will be less noisy than a Shuttle take off. That still leaves a lot of room for it being *much* noisier than *any* conventional airport would accept, *even* with a near empty Hydrogen only ferry flight.

[93143]

I'd thought SABRE in air breathing mode was running at around 100 atm c1500psia.

I didn't check; I was just trying to ballpark some numbers in a hurry.

Sea level optimal expansion drops to ~2200 m/s.  Altitude result stays about the same.

Incidentally did that run give any idea what the plume size would be?

I don't think CEA does that.

It would be underexpanded at sea level without one of those advanced nozzles.  With one, it should roughly match ambient pressure up to the point where the expanding flow fills the nozzle, after which it should start to expand laterally after exiting the nozzle.  I could probably try to compute this effect, but it would be rather involved and right now I need to sleep...

[adrianwyard]

Incidentally did that run give any idea what the plume size would be? People has expressed concern over plum impact on Skylons tail. I think this is highly unlikely given the engine locations on Skylon *unless* the plume is huge.

I should admit the two plume cases I mentioned were simply based on eye-balling it. In the first case, at rotation when it first becomes airborne, the engines look far enough ahead of the tail that the plume could reflect off the runway and cause heating back there.

And the second case - at vacuum - is simply based on noticing how very wide the plumes expands from viewing rocket-based cameras. But since there's no way to judge scale it's quite possible the apparently wide plume is a great distance behind the rocket...

It could be a non-issue.

Constraint of the plume would come along for free if they do use altitude compensating nozzles, such as the expansion/deflection type they've looked at.

[lkm]

New Alan Bond interview from Farnborough talking about the precooler test program

http://www.rocketeers.co.uk/node/2370

[john smith 19]

New Alan Bond interview from Farnborough talking about the precooler test program
http://www.rocketeers.co.uk/node/2370

Fascinating.

Looks like someone did have their phone out at Farmborough 

This is very good news. The frost control system looks very good, leaving the full temperature range as the only open area.

This must be close to getting their next round of funding released on this basis.

So I guess we're looking at press release or site update in a couple of months. 

[flymetothemoon]

So I guess we're looking at press release or site update in a couple of months. 

Sadly I did hear someone say during Farnborough (was it that last Mr Hempsell interview?) that the next phase of the test programme would be running to the end of the year, so I am kinda thinking that we won't be hearing much before then...

[PMN1]

BBC 4 10pm Wednesday 12th September

http://www.bbc.co.uk/programmes/b01mqv45

For his entire life, one man has nursed the dream of putting mankind into space. Inspired by the Dan Dare comic strip, Alan Bond first started building rockets as a teenager in his back garden. He started his career working on Britain's Blue Streak rocket, then HOTOL - the world's first attempt to build a 'single-stage-to-orbit' spacecraft. Each time, he was thwarted by lack of funding from the UK government, so, together with two colleagues, Richard Varvill and John Scott-Scott, he decided to go it alone. This documentary tells the story of how the three rocketeers defeated the Official Secrets Act, shrugged off government intransigence and defied all conventional wisdom to build a revolutionary new spacecraft - Skylon