Bjorn’s Corner: Supersonic transport revival, Part 12

October 26, 2018, ©. Leeham News: In the previous Corner we discussed the noise challenge an SST engine has. To be effective at Supersonic speed it needs a high Specific Thrust (a fast jet out the back) but this creates Takeoff and Landing noise.

We now look at some key data for SST engines.

Figure 1. The GE Affinity medium bypass Turbofan for Aerion AS2. Source: GE.

SST engines

The first SST engine, the Concorde Olympus was designed in the 1960s, almost 60 years ago. One would think it should be easy to design an engine for SuperSonic Transport (SST) aircraft which would be more efficient than the Rolls-Royce/Snecma Olympus 593.

By today’s standards, its data is antiquated. It’s a straight jet engine with an Overall Pressure Ratio (OPR) of 15.5 and with a cruise specific fuel consumption at Mach 2 of 1.2 lb/lbf/hr. Modern Turbofans are below 0.55 lb/lbf/hr, less than half the Olympus.

As we shall see it’s hard to beat the Olympus for a Mach 2 engine. The reason is the Olympus combines a low mass flow at cruise of less than 90kg/s or 200lbm/s with a high specific thrust of more than 500m/s or 1,600ft/s (the 900m/s value we talked about in the previous Corner was for Takeoff with Afterburner). The low mass flow gives a low Ram drag, which compensates for other inefficiencies.

But a straight jet is impossible to use with today’s noise regulations. We must use bypass turbofans to reduce takeoff and landing noise.

The Mach 1.4 engine

Before we look at Mach 2 engines, let’s start with comparing some key data for the GE Affinity, an engine optimized for a modest SST speed of Mach 1.4. We compare it to its donor engine, the CFM56.

The CFM56-5 sitting on an A320 has a cruise thrust of around 5,000lbf at 37,000ft, a bypass ratio of 5 and a fuel consumption of 0.6 lb/lbf/hr.

To produce this thrust the engine produces a Gross thrust of 12,500lbf. Gross thrust minus Net thrust gives us our Ram drag. It means we have a Ram drag at Mach 0.78 of 7,500lbf and we have a Gross thrust to Net thrust ratio of 2.5.

We now compare these values with the values for Affinity at Mach 1.4. To keep Ram drag within bounds, GE set the ByPass Ratio (BPR) of Affinity to around 3.

If the engine shall produce 3,500lbf thrust per engine at Mach 1.4 and 50,000ft we need to inject fuel in the combustor to produce 11,000lbf of Gross thrust. Our Gross thrust to Net thrust ratio is now 3.15.

The Ram drag increase for the Affinity is reasonable. This has been achieved by keeping the Bypass ratio down at ~3 instead of the original engines 5. The cost is a loss in propulsive efficiency.

The increased Ram drag and the lower propulsive efficiency increases the cruise fuel consumption of the Affinity at Mach 1.4 with about 50% over the CFM56-5 at Mach 0.78.

Mach 2 engines

In the next Corner, we will look at engines for SSTs flying at Mach 2 or faster.

22 Comments on “Bjorn’s Corner: Supersonic transport revival, Part 12

  1. Tip: There are some good discussions re the Affinity in the comment section for Part 11 for those interested. We can continue the discussion here.

  2. Is the fuel consumption really only 50% higher at almost twice the speed? That would mean a lower total fuel consumption per mile! (Besides the increased drag of the plane, of course).

    • I remember reading some place that, since the concorde was around Mach 2, it was indeed one of the most reasonable in fuel consumption per mile for the era. the problem was that during axi and take off, it consumed so much fuel, that in some instances, it has to get back to gate and reload fuel because of a delay in departure. So, yes, for an aircraft that is designed to cruise supersonic, it is always better to say as long as possible in supersonic speeds.

  3. I’m puzzled by the same thing here: “with a cruise specific fuel consumption at Mach 2 of 1.2 lb/lbf and hour. Modern Turbofans are below 0.55 lb/lbf and hour, less than half the Olympus.”

    Wouldn’t that imply that the Concorde has roughly the same fuel burn as a modern jet, given it’s more than twice as fast?

    • No, because you’re not considering the abysmal lift to drag ratio of supersonic flight. So while the specific fuel consumption is impressive, multiply it by thrust required (gives actual flow rate), divide by the number of passengers, and multiply by cruise time. Then you’ll have the fuel used per passenger for the cruise segment and I doubt it will look efficient.

    • Looking at an extensive table of consumption per seat mile on Wikipedia, the Concorde used 5 times more fuel per passenger mile than current subsonic quads.

    • Remember the Concorde had four engines and carried about 100 people, compared to a modern widebody that has two engines and carries 300-odd passengers.

      So fuel burn per second per engine may be similar to a modern widebody, but with twice as many engines and 1/3 as many passengers, fuel burn per passenger km will be something like six times higher.

  4. given the clutched lift fan design used in the F135 on the F-35B (PW/RR tech?) it would seem to me that it would be feasible to have a clutched high bypass fan (think a GTF with a clutch instead of a gearbox or maybe even a clutch and gearbox) and intake doors to create an engine that is a high bypass TF at takeoff and a high specific thrust and low ram drag at cruise.

    you would pay a penalty in terms of cross sectional area, but gain enormous operational flexibility of GTF like subsonic fuel efficiency and noise profile while having the desired attributes for supersonic cruise (and a modern core)

    another possibility would be to have discrete LP and HP systems on different centerlines with airflow controlled by doors such that the LP system is completely shut down at high speed cruise and the HP core is naturally operating at optimal mass and pressure ratios…. I believe there was speculation that the “sr-72” was based on this model

  5. Your units of fuel consumption are incorrect.
    They should be lbm/lbf an hour, not “and hour”

    • If we’re going to use proper symbols (with no subscripts available in plain text), it ought be “lb/lbf/h”, no?

      (Let’s be thankful there is no appearance of “nm” for nautical mile!)

      • Seeing nm is my favorite… flights tend to go more than the width of a couple human hairs.

        Also agreed with lb/lbf/h (or lb/lbf-h) instead of lb/lbf and hour… I’ve never seen “and” used for unit notation. Good article/series, though.

  6. Björn, I like to read your posts, also for their technical quality. But I’m sorry to say that several things mentioned here don’t make sense to me, and also I can’t see the point of the overall post.
    For sure the Olympus was a very good engine, yet it simply can’t be true that it is not possible today to build a significantly more fuel efficient engine. Actually, you confirm this by concluding that the Affinity ar M1.4 might land at 50% higher sfc than the CFM, so 0,9, which is actually 25% better than the Olympus. Ok, the Olympus went M2 versus M1.4 of the Affinity, but, seriously, the CFM HPC is a piece of junk from the 70s/80s and far from representing current core engine technology. Further, going faster does not only hurt by Ram Drag, the engine also benefits from ram compression.
    Reducing the BPR from 3 to 5 might sound serious to average Joe, maybe biased by Leahys ‘Bigger is better’ mantra, yet by the numbers it increases fuel burn by less than 10%, so no big deal. Note that the Leap X needed to go from 5 to 12 AND almost double core OPR significantly to “just” gain less than 15% in sfc. Why should it be so much different when going the other direction?
    The only reason why the Aerion has the Affinity is not because the Affinity is a good or the right engine, but simply because no one else woukd do them a better engine – which would be perfectly doable, by far, but would never pay off, as there is no serious SST business case.
    For the Olympus, the best designers of their time gave it all to do what ever was possible. The Affinity is the lowest risk BC engine that would just fulfill the minimum requirements. An outdated core with fan and nozzle chosen to just make it work.

    • Even the Olympus 593, the Concorde engine was developed from the earlier Olympus versions, which were subsonic and flew in the early 1950s. The Concorde first flight was in 1969.
      So you can see even then they used an existing engine as stepping stones, the final proposed version of the Olympus , the 662 would have eliminated the afterburner and made a big difference for takeoff noise and improved range but that wasnt to be.
      I think Bjorn has explained why some of the newer engines wouldnt be suitable as a core for this plane, which of course wont be made in 1000’s per year.

      • I have vague recollections that the TSR2 programme also did a lot of the technical development required to get the Olympus to go supersonic.

        Getting up to speed and staying there did good things for mileage. The SR71 would become more efficient the faster it went. Kind of a trade-off between fuel consumption and melting!

        • The SR71 engine as previously mentioned was semi ram jet and at cruise used the after burner continuously, but at Much 3 plus you have a lot more issues to solve

    • Hi Karl,

      the point is if we could design a straight jet or very low bypass turbofan engine for a Mach 2.2 SST and get it to pass Stage 5 noise regulations it would be easy to make an engine which is more efficient than the Concorde’s engine. But we can’t! This is the point. The efficiency is measured in TSFC, Engine diameter (because of volume wave drag) and weight.

      Present SST engines have to be designed sub-optimal due to modern noise regulations, the more the faster they need to cruise. It’s what we discussed in this series and will be very clear when we look at M2.2 engines. (If you then don’t believe me get a free trial of Gasturb (, used by the engine OEMs) and try to make a noise acceptable engine which is more efficient than the Olympus at Mach 2.)

      You need to go to a variable cycle engine to crack this nut and these are not available yet.

      • If you got “help” during T-O with a catapult up to 250kn the numbers will change for a supersonic engine both regarding noice, T-O fuel consumption and airport emissions. To get FAA/EASA certify it is another problem. The pax can handle maybe 5g’s for some time (like max sustained g’s at amusement parks) , do mot know the Disneyland limit…

        • Part of the reason why catapults used on US carriers was just to shorten the takeoff run. The down side to that was you cant rotate the plane ( for higher lift from better angle of attack) before you leave the deck. The solution to that was the ski jump.
          The rest of the idea of catapult for a supersonic business jet doesnt seem well thought. You dont have a short takeoff distance and you would still need high power setting at takeoff ( it doesnt have a noisy AB) and as Bjorn has pointed out previously the critical thrust level is still that required at cruise altitude to push past the sound barrier.

          • In this performance/noise study you can use land catapults differently as they can allow a longer T-O run than the limited runway on an aircraft carrier, the engine thrust can be way down just to be able to spool up to climb thrust just before rotation to reduce noice and fuel burn, as the spool up time is pretty quick on todys FADEC equipped engines. If you maximize the reduction in noice and fuel you catapult to a speed way buyond Vr (as fast as the tyres and carbon brakes can handle). The Do-Dodonpa on Fuji-Q Highland amusement park does 0 – 111.9 mph in 1.56 seconds (3.27 g)…

    • ‘ but, seriously, the CFM HPC is a piece of junk from the 70s/80s and far from representing current core engine technology.’

      I believe the CFM56 HPC and turbine were derived from the GE F101, i.e. B–1A/B engine…hardly a ‘piece of junk’. It was sophisticated enough that the Carter administration only reluctantly allowed the CFM consortium to use it in the CFM-56 series.

  7. Those fixed stators could cause troublesome fan blade vibration and also noise for the new highly loaded GE engine. Refer to ” Tyler and Sofrin” pioneering paper from P&WA.

Leave a Reply

Your email address will not be published. Required fields are marked *