AFRL Review Finds Skylon’s SABRE Engine Theoretically Viable

Skylon with the SABRE engine. (Credit: Reaction Engines)
Skylon with the SABRE engine. (Credit: Reaction Engines)

OXFORDSHIRE, England (Reaction Engines PR) — Reaction Engines Ltd. is pleased to announce that analysis undertaken by the United States’ Air Force Research Laboratory (‘AFRL’) has confirmed the feasibility of the Synergetic Air-Breathing Rocket Engine (‘SABRE’) engine cycle concept.

The analysis was undertaken by AFRL as part of a Cooperative Research and Development Agreement (‘CRADA’) with the Air Force Research Laboratory’s Aerospace Systems Directorate (AFRL/RQ). These investigations examined the thermodynamic cycle of the SABRE concept and found no significant barrier to its theoretical viability provided the engine component and integration challenges are met.

Reaction Engines Ltd. and AFRL are now formulating plans for continued collaboration on the SABRE engine; the proposed work will include investigation of vehicle concepts based on a SABRE derived propulsion system, testing of SABRE engine components and exploration of defence applications for Reaction Engines’ heat exchanger technologies.

AFRL/RQ program manager Barry Hellman stated – “The activities under the CRADA have allowed AFRL to understand the SABRE engine concept, its pre-cooler heat exchanger technology, and its cycle in more detail. Our analysis has confirmed the feasibility and potential performance of the SABRE engine cycle. While development of the SABRE represents a substantial engineering challenge, the engine cycle is a very innovative approach and warrants further investigation. The question to answer next is what benefit the SABRE could bring to high speed aerospace vehicles compared to other propulsion systems. Although application of the SABRE for single stage to orbit space access remains technically very risky as a first application, the SABRE may provide some unique advantages in more manageable two stage to orbit configurations. Furthermore, the heat exchanger technology also warrants further investigation for applications across the aerospace domain.”

Sam Hutchison, Director of Corporate Development at Reaction Engines Ltd commented – “The confirmation by AFRL of the feasibility of the SABRE engine cycle has further validated our team’s own assessment and conviction that the SABRE engine represents a potential breakthrough in propulsion that could lead to game changing space access and high speed flight capability. We look forward to continued collaboration with AFRL”.

SABRE is an innovative class of aerospace propulsion that has the potential to provide efficient air- breathing thrust from standstill on the runway to speeds above Mach 5 (4,500mph) in the atmosphere – twice as fast as jet engines. The SABRE engine can then transition to a rocket mode of operation for flight at higher Mach numbers and space flight. Through its ability to ‘breathe’ air from the atmosphere, SABRE offers a significant reduction in propellant consumption compared to conventional rocket engines which have to carry their own oxygen – which is heavy. The weight saved by carrying less oxygen can be used to increase the capability of launch vehicles including options for high performance reusable launch vehicles with increased operational flexibility, such as horizontal take-off and landing. Additionally, the SABRE engine concept could potentially be configured to efficiently power aircraft flying at high supersonic and hypersonic speeds.

  • windbourne

    While this will not be able to compete against spacex, I hope that it is built. The ability to put small loads several times a day, is staggering and needed.

  • Douglas Messier

    I dunno. This report is basically saying it’s theoretically feasible of they can actually build and integrate all the components. Well, yeah…. And if you keep paying us, we can help you figure out how to do that.

    Less here than meets the eye, maybe.

  • Geoff T

    Whilst I don’t necessarily disagree with you on that assertion. Assuming Skylon was actually developed into a working space plane how would it be at a disadvantage to SpaceX? An aircraft form factor would be significantly easier to service, refit and refly than a conventional stack rocket in theory (compare taxiing to a hangar to reassembling the separate stages of a Falcon 9 and getting it back on the pad). Payload to LEO seems comparable too.

    The only firm advantage I can see is that SpaceX is here and flying today, whilst Skylon is on paper with a lot of unknowns ahead. Of course, that’s a pretty solid advantage in and of itself!

  • Doug Weathers

    I can’t figure out why a space vehicle would want to loiter in the atmosphere any longer than necessary. Get out as soon as you can!

  • Valerij Gilinskij

    There is a limit starting mass flying up horizontally, on the wings of the machine. Therefore, there is a limitation of capacity. In addition, the likely service Skylon between flights is not easy and takes a lot of time.

  • Hug Doug

    It wouldn’t; takeoff aside, it would largely launch much like a rocket normally does. However, during the time that it is in dense atmosphere, it can use the oxygen in the air for oxidizer instead of carrying it in an onboard tank.

  • windbourne

    Costs. Basically, spacex will have all of their R&D paid for before skylon hits runway.

  • Hug Doug

    Major disadvantage is development costs and payload the rocket can put into orbit.

    Skylon’s development costs are estimated at about $10 billion. Payload to orbit is about 33,000 lb.

    Any way you analyze it, Skylon’s amortized costs per launch compared to mass of payload put in orbit turn out to be worse than SpaceX’s. About the only way to avoid that is if SpaceX falls flat on its face financially and goes bankrupt (granted, there are still some people who think that will happen).

  • Chad Overton

    Seems like a great thing for the UK and USAF to jointly fund development and buy down some of the risk.

  • Solartear

    Flying to 1/5 of orbital speed without using any of the liquid oxygen helps a lot. And it can get well over 3,500 Isp while in atmosphere.

  • stoffer

    You win Isp, but put on heavy engines. I am not sure of the trade-off. It may make a great intercontinental plane or even a launch assist, but an SSTO?

    I would rather take Philip Bono’s Rombus-like thing and add a first stage to push it up to say 100 km altitude to shave the gravity and air drag of the second stage. With hydrolox 7 km/s can be reasonably done in one stage.

  • mzungu

    With the state of UK’s aero industry, and without a US partner, not going to happen… With the struggle just getting the Ariane 6 Powerpoint slides together, don’t think they be getting any help from the French or the Germans any time soon.

  • windbourne

    My thinking is that skylon will be too little, too late, unless they push the airline model. Create new destinations which is serviced only by skylon.
    Itherwise, spacex has 20-30 launches per rocket, they will be much lower costs than skylon due having dealt with R&D costs over the next 10 years.

  • windbourne

    In addition, it will likely spend a bit more time in 75-100k ‘ range to move from subsonic, up to say mach 5-10 prior to switching to rocket.

  • Michael Vaicaitis

    Don’ compare Skylon with F9. Should Skylon ever be made, it won’t be before 2025-2030, by which time SpaceX would hope to have BFR (150-200 tonnes to LEO ???). Even if they also have a medium lift rocket, it won’t be the kerolox F9. Also, Skylon will cost $750 million each to build and only have a lift of 15 tonnes. Beautiful as Skylon would be, the numbers don’t stack up in it’s favour.

  • Doug Weathers

    I’m familiar with the argument. I just don’t think the tradeoffs make sense. In order to pull oxygen out of the air, you need to stick machinery into the airstream. This automatically increases drag, proportional to the amount of oxygen you are extracting.

    Also, almost all of the oxygen is right near the ground. You’re not going to spend much time there.

    Now add in the weight and complexity of the oxygen extracting hardware, and the sheer size of the vehicle (it’s longer and heavier than an Atlas V, and carries less payload), and it really starts to look silly to me.

    Skylon is a huge jump in technology past where we are now. Trying to go too far in one jump has gotten us in trouble in the past. Staging is not as scary as we used to think – if it did nothing else, SpaceShipOne and WhiteKnight showed us that.

    Of course, the above is only my opinion. I haven’t done any trade studies, and I know that a lot of dedicated people have been spending the last decade or so working on it. I wish Reaction Engines the best of luck.

  • DavidR2015

    Each Skylon is only expected to last around 200 flights, unlike airliners.

  • DavidR2015

    I’m wondering if, when all said and done, if REL will be able to find enough competent sub-contractors in Europe to be able to build this machine. It may need US industrial expertise and personnel to turn the design into reality.

  • Duncan Law-Green

    “The importance of the aerospace industry to the economy has changed noticeably since 1997. While output within manufacturing fell by 4.5% between 1997 and 2013, output in the aerospace industry increased. Aerospace repair and maintenance increased by 8.6% while output for aerospace manufacturing increased by 79.6% (Figure 1)5.”

    Source: Office of National Statistics: http://www.ons.gov.uk/ons/rel/uncategorised/summary/changing-shape-of-uk-manufacturing—aerospace/sty-uk-aerospace-industry.html

  • Michael Vaicaitis

    Yes, 200 flights is their stated aim – of course reality may yield only 100 or it may yield 500. I have a sneaking suspicion that Alan Bond would want to be conservative about longevity, so I tend to think of 200 as a minimum. Also, it’s not clear whether that estimate of 200 was motivated by the SABRE engines or the Skylon airframe.
    In a similar vein, Elon Musk has spoken about reusable boosters lasting 1000 launches. Obviously not F9, but perhaps BFR or some later methalox architecture. Even if we treat the figure of 1000 as potentially fanciful, it does give us at least some insight that SpaceX thinks somewhere near that might be possible. Also, the fact that both Musk and Bond are both talking about reuse numbers in the same ballpark, gives some credence to both of their claims.

    Last I read, the one-off Skylon cost was to be about $650M and the production line cost may go down to $450M. I took the liberty of extrapolating up to $750M, both to allow for any optimism in those guessimates and to make the point that it will be expensive. As a counterpoint, back in the day before SLS, Musk made the offer of a 150 tonne SHLV for a development cost of $2.5 billion and a launch cost (presumably expendable) of $300 million. From that we might expect that BFR will also be around those costs too. The problem for Skylon though is that BFR will still likely be cheaper, in development, manufacturing and flight costs, but also it’s 15 tonnes to LEO versus 150+ tonnes. It would be interesting to see SABRE on the first stage of a two stage vehicle, but I don’t see how Skylon, or any SSTO, can compete on any meaningful metric.

  • Michael Vaicaitis

    I don’t think there is the expectation that Skylon could launch “several times a day”. Alternatively, you could launch a reusable two stage rockets from various worldwide locations everyday and achieve the same end – with the added benefits of reduced costs and reduced cost risk on any given launch. SABRE is feasible (if you can handle the LH2 and Helium), but Skylon is just pretty.

  • Michael Vaicaitis

    I think the Skylon airframe is relatively straightforward manufacturing issue compared to the SABRE engines. The heat exchangers are the difficult part of manufacturing the engines and they’ve solved that problem already. I think there is the required manufacturing expertise in Europe for test articles and one-offs. If it ever reached production line status, Airbus for one, should be quite up to the task.

  • therealdmt

    Jackpot

  • Hug Doug

    IIRC it saves a few hundred tons of weight in oxidizer.

    Actually, the only tech that had big question marks associated with it were the pre-coolers, and those were validated a couple of years ago. Most of the rest was developed for the HOTOL in the 80s. So it’s not as big a jump as you make it out to be. That said, it does need to be developed into technological maturation, that takes time and testing.

    It’s great technology, but I think that when SpaceX starts reusing stages and their prices drop accordingly, Skylon will be quietly shelved.

  • Solartear

    As 2-state-to-orbit it would cost an expensive, large second stage for every flight, hurting the economics and not making Skylon concept more expensive.

    Reaction Engines has had outside organizations, including ESA, verify their numbers make sense as SSTO.

    People have initial trouble conceptualizing Skylon because they think of SSTO like DeltaClipper or StarVenture where they need extreme margins. USA tried to do launch concepts using air (LACE,etc) decades ago, but gave up because the air-coolers would freeze. SABRE solved this problem.

    Specifically regarding heavy engines, their is not much mass added since it is basically just extremely light cooler plus compressor stuck onto front of regular rocket engine. The wings seem like extra mass to me, but they have done the math, and had it checked, to show it makes sense.

  • mzungu

    I was thinking more in longer terms that dated back to the 80’s or even earlier…. no doub it gotten better. I just don’t think UK or any of it’s companies have the will to design and build a whole airframe anymore…. or even the engine. for that, it needs partnership that it prob won’t find in Europe.

  • Duncan Law-Green

    “it’s companies have the will to design and build a whole airframe anymore…. or even the engine”

    Err.. you have heard of http://www.rolls-royce.com/ , right? A UK company and one of the largest aero engine manufacturers in the world? REL just hired their Chief Engineer for Technology and Future Programmes. With RR support, building a working SABRE engine is well within their capabilities.

    You can argue that the UK doesn’t have experience in completing large airframes any more, but *nobody* has experience in building the kind of airframe that the Skylon is — it’s closer to an airship spaceframe structure than a conventional aircraft fuselage.

  • windbourne

    “USA tried to do launch concepts using air (LACE,etc) decades ago, but gave up because the air-coolers would freeze. SABRE solved this problem.”

    Not just that. Back in the 80’s, the materials were NOT ready to handle this. Since the work on NASP, a lot of R&D has gone into materials that have allowed for SABRE to work.
    But, yeah, it seems like SABRE is the last R&D to make it work.
    However, I still suspect that R&D costs on making this work, will mean that it will become the shuttle.

  • windbourne

    That is exactly what I am afraid of.
    However, in one positive way, it will be the start of SSTO.

  • windbourne

    Actually, musk has said that F9 is designed with 40% safety factors, rather than 25% as is normal.
    As such, he expects that it will last 100+ times for the frame.
    The engines are a different issue.

  • windbourne

    In fact, there are loads of ppl HOPING that Musk and/or SpaceX will fail.

  • Michael Vaicaitis

    The structure of a rocket will experience more stress than Skylon is designed for. Although a “simple” cylinder is also a fundamentally more robust design. F9 specifically has not optimally designed in the first place, since it’s primary design parameter is that it should be transportable by road, hence it is long and thin. For F9 the kerosene engines are surely the weak point for longevity.

  • Terry Jagers

    Regarding the air coolers I read somewhere that at high mach its cooling the air by 1,800 deg in just a few seconds which is a huge breakthrough

  • Gath Gealaich

    The RS-25 was also theoretically viable, and even got eventually built, it “merely” turned out to be uneconomical to operate because of its extreme parameters.

  • Gath Gealaich

    “It would be interesting to see SABRE on the first stage of a two stage vehicle, but I don’t see how Skylon, or any SSTO, can compete on any meaningful metric.”

    There was a study that compared air-breathing engines with the option of using PDRE engines, and PDRE engines won. You also have to wonder how having an 450 s vacuum Isp SABRE is beneficial compared to having a 490 s vacuum Isp PDRE whenever you want to get beyond low Earth orbit.

  • Gath Gealaich

    “For F9 the kerosene engines are surely the weak point for longevity.”

    But surely simple kerosene engines like the Merlin 1D can’t be a *weaker* point for longevity than complicated hydrogen engines like the RS-25 or the even more complex SABRE? We’ve learned so much from our real-world experience so far.

  • Gath Gealaich

    “As 2-state-to-orbit it would cost an expensive, large second stage for every flight,”

    Yes, and an SSTO requires an even more expensive single stage for every flight, just with a worse payload fraction. So there’s not much of real difference there.

  • Solartear

    No, Skylon is designed to be reused hundreds of times with very little maintenance, in stark contrast to launchers with expended first stages and the high-maintenance STS program.

    Plus the technologies of rocketry often do not scale well. How are you going make the Skylon a lot cheaper by making it 2-stage-to-orbit(2STO)? The SABRE engines would be more expensive to scale down, running into problems like STS had.

    The basic shape of Skylon would have to remain the same as 2STO for aerodynamics and to contain the large second stage needed.

    The payload fraction does not work the same when the air is used for a large portion of the launch. Traditional rockets have to carry all their oxygen, so they need higher margins to get the same payload to orbit. So while the payload fraction is a problem for a SSTO like VentureStar, it is not for Skylon.

  • Gath Gealaich

    1) There’s no technical requirement for a TSTO vehicle to be expendable.

    2) I have no doubt that in the wet dreams of its designers, Skylon is being reused for hundreds of times without maintenance, but one of the major reasons for STS’s high maintenance costs were the technologically extreme engines that Skylon’s engine design actually outdid in complexity. (BTW, what is the contingency scenario for in-flight engine failure on Skylon? Those highly-off-axis units certainly don’t make me feel very safe.)

    3) I find amusing the claim that “the technologies of rocketry don’t scale well” when Skylon is limited to 15-tonne payloads whereas classic launch vehicles can fairly easily scale to 150 mt payloads and large payload volumes. Once you get to the point of assembling spacecrafts in the multi-hundred-tonnes range, the design problems, logistics costs, and added mass of all those mating mechanisms will quite severely reduce the savings you’d accomplish with even reasonably cheap Skylon flights. To make that clearer, imagine having a launcher that can launch half-tonne payloads completely for free. Now design and implement a manned Martian mission with it.

    4) A TSTO vehicle obviously wouldn’t use SABRE engines. That would be useless weight; either hydrogen-oxygen PDREs, or even better, tripropellant methane-hydrogen-oxygen (variable Isp) PDREs would be used for the orbiter. Those could easily have much higher top Isp (in the 470-500 s region) – and even much higher average impulse density in the tripropellant version, making the orbiter smaller – all while being vastly simpler in design than SABREs, and perhaps even having significant redundancy due to the number of impulse tubes.

    4) Skylon needs to cover about 80%-85% of its delta V or something like that with its on-board oxygen. Thus, the claim (regarding Skylon, I presume?) that “air is used for a large portion of the launch” is quite nonsensical. With Skylon, air is actually used for a very small portion of the launch.

    5) Payload fraction is meaningless anyway, all that matter is simplicity, reliability, and economy. Basically, cost.

  • Bigdog

    Yes the world is full of idiots isn’t it?

  • Bigdog

    Fractions of a second.

  • Bigdog

    “Staging is not as scary as we used to think – if it did nothing else, SpaceShipOne and WhiteKnight showed us that.”

    What? SS1 is suborbital. Its not a space craft at all.
    It has nothing to do with “staging” at all.
    Apples and oranges pal.

  • windbourne

    Not sure that idiots is the right word as much as egotistical or simply jealous.

  • Doug Weathers

    SpaceShipOne, as its name implies, was a ship that went into space. Really. It actually got there! Specifically, this means that it was a vehicle with a person inside, and it went higher than 100 km above the surface of the earth.

    Yes, it was suborbital. So is a sounding rocket, and so was Alan Shepard’s first flight. Doesn’t mean he didn’t make it into space.

    There is more than one definition of “space”, BTW. They all have to do with altitude. The one that most people use these days is the 100km/62mi limit, sometimes called the Von Karmann Line. The USAF used to use 50 miles.

    As for staging, yes, WhiteKnight was the first stage of a two-stage space launch system. It was also an airplane, but that’s not important right now.

  • Toby Lin

    Not just oxidiser – it also allows the vehicle to use something like a thousand tonnes of atmospheric nitrogen as reaction mass. This slows down the exhaust velocity but massively increases the airbreathing Isp to something like 3500s!

    Think of the Atmosphere as the 1st stage drop tank!