Sunday, October 1, 2017

Another A380 Loses an Engine in Mid-Flight

By now, everyone has heard about the Air France A380 flight yesterday that had an engine break up in mid-flight over the North Atlantic (one report says over Greenland).  The aircraft declared an inflight emergency, was re-routed from its path to Los Angeles and landed safely at Goose Bay airport in Canada.  The pilots handled the emergency well and the plane landed without further problems. 
Passenger photo.

The thing is, this isn't the first engine issue they've had, on an airframe that hasn't flown that long.  Soon after the A380s were delivered to Qantas airlines in Australia, 2010, one of their A380s lost its port side inboard engine to an explosion on takeoff.  The engine threw debris through the wing of the aircraft and could have just as easily thrown debris through passengers.  The pilots handled the emergency expertly and safely landed the aircraft.  The cause was judged to be a quality problem at Rolls Royce Trent engines, and resulted in grounding of several aircraft, including among different airlines to have engines inspected or replaced. 

Are two midair engine explosions in seven years unusual?  That's a little awkward to answer because if you search for any model aircraft, you'll find stories about engine explosions.  Despite that, I'm sure that everyone in the design chain does their best to make sure they don't explode.  Is it worth asking if the A380 system has some issues with its engines?  Not so much that the engines aren't built right, which was the case with the Qantas incident, but if there's something about their operating envelope that wasn't adequately specified?  What I mean is that the engines were specified, designed and then tested to a specific set of requirements, so I'm sure they meet the requirements.  Is it possible that the requirements are wrong? 

It's rough for outsiders to the industry to find real numbers for the number of flight hours, failures per flight hour and so on.  We always have to keep in mind that if two very rare occurrences happen in rapid succession, they might legitimately still be very rare.  People do hit the lotto twice.  It does, however, make me want to keep an eye on the A380 from a respectful distance. 


23 comments:

  1. Two in seven years is not unusual but you definitely have to determine the cause as it could be a manufacturers assembly problem, perhaps a design problem or maybe change inspection intervals to shorter times. Was this an engine problem or a engine cowl problem, was the engine temp or pressure exceeded, had any turbine blades been damaged and blended improperly. It will likely take a while to figure out the actual problem. Now my personal preference is to fly Boeing as they have a good flexible wing for a smooth ride. Airbus has a stiff wing and reminds me a driving down a washboard gravel road. indyjonesouthere

    ReplyDelete
  2. Sometimes things look chaotic because there isn't enough data on what happened. I think that's your point.

    ReplyDelete
    Replies
    1. The statistical question that's all around these days is something like, "how can we have two 100 year storms in just a few years?" That's entirely possible; statistics are only for the long run. It doesn't mean the storms come exactly every 100 years. It tells you what the probability of two in rapid succession is, but it doesn't say it's impossible.

      Two engines with a million hour MTBF could fail in a thousand hours and really have a million hour MTBF. (I don't know what the actual, rated MTBF is, or the number of flight hours)

      But it makes you wonder if that MTBF is accurate. It makes you wonder if the models are correct.

      A friend points out there's a video (this one, I think) of a test of the Rolls Royce A380 engine where they blow off a turbine blade to verify that the nacelle contains it all. It contains the debris. That didn't happen in this latest incident.

      Delete
    2. I have see the video, and yes, the nacelle contains the catastrophic failure. But it was in static conditions. It didn't have a 500+ MPH wind to keep it spinning. One of the issues is trying to replicate actual in-flight conditions.

      Delete
    3. One of the issues is trying to replicate actual in-flight conditions. Exactly. Passing the test doesn't matter much if the test is unrealistically easy.

      One of the standard tricks is to say since we don't know exactly how well the test corresponds to reality, we're going to test it at much higher stresses than we expect. Maybe it still isn't a tough enough test.

      Delete
  3. Does the A380 use an engine that was designed specifically for the A380? Or does it use an "off the shelf" design that other airliners use. Instead of looking for engine blowouts on the A380 you should find out the type of engine used and do a blowout search for that specific engine. I don't know if the engines are Rolls-Royce, GE or P&W's, but it just makes sense to design engines for use in as many different plane types as possible.

    ReplyDelete
  4. Missed the link you included to the RR Trent 90 website. From looking at the site, you can tell that they're proud that the A380 "chose the Trent 90". But this is obviously a sale pitch page. So...how many Trent 90's are there out there and how many have blown?
    Delivering the lowest lifetime fuel burn and excellent environmental attributes, it’s easy to see why the Trent 900 is the engine of choice for the Airbus A380.

    ReplyDelete
    Replies
    1. I think that pretty much all new aircraft get new design engines. Introducing a new airframe, especially one the size of the A380, is a rare thing. The major air frame makers may do a new jet per decade; there has never been one the size of the A380, and the next biggest, the 747, has been flying since the 1960s. They're not going to use a perceived lower tech engine when the engine makers are continually saying they can improve efficiency. Airlines live or die by their fleet fuel usage.

      Because of that, I'd bet that RR designed that engine specifically for the 380. Which doesn't mean they can't have more than one supplier. It's not that uncommon to have more than one supplier of similar, but not identical engines.

      Delete
    2. I think you kind of have it backwards. You're right that no one would expect Airbus to use an older engine design. But you seem to assume that the old airplanes would continue to use the old engine designs. They don't. As more modern, energy efficient engines come out the only question becomes, will the new engine fit under the wings. And, somewhere I saw an article saying that one of the Middle East airlines was considering changing their engines to the RR Trent 90's on their A380's. They had chosen another brand engine and they weren't happy with them. Everything on a modern airline is optional: engines, cockpit setup, seating, cargo capacity vs seating and so on. If you'd like I'll try to find it again...

      Delete
    3. Trent 900's. I don't know why I keep typing "90's"

      Delete
  5. This failure was "contained". The engine failed but the parts did not penetrate the wing, thus not taking out hydraulics or the fuel tank. Apparently this is the second engine failure in ten years. Of course we would like to have zero engine failures but that is not realistic. Jet engines are incredibly reliable but they do fail occasionally.

    ReplyDelete
  6. Doesn't look contained to me. The whole Fan stage is missing; note the sheared hub. It, or segments of it, departed the aircraft on paths which did not impinge the inboard engine, wing or fuselage, which account for about 25/360 degrees(7%)of possible trajectories. They got lucky.
    Differ

    ReplyDelete
    Replies
    1. I'm with you, Differ. They got lucky.

      Compare the engine pylon on the inboard engine to the outboard. Everything forward of the start of the taper down to the nacelle is gone. A lot of the engine left the scene of the accident and the fact that it didn't go up into the wing is partly because it was too far forward of the wing. The Qantas incident did throw metal through the wing.

      I don't see any way that getting lucky isn't good. Ragman is correct that jet engines are incredibly reliable, but do fail on occasion. When they do fail, you want all the luck you can get.



      Delete
  7. Shhheet, The BAE-146 used to shuck engine parts out of the Lycoming ALF -502, like no other!.
    I remember a PAX had a turbine blade lodged in his briefcase that had been in the overhead storage compartment.

    ReplyDelete
    Replies
    1. Yikes! That scored a 10 on the pucker meter.

      Delete
    2. BAE=Bring Another Engine. The A380 engine parts did not penetrate the wing or the cabin. This is exactly what it is designed to prevent.

      Delete
  8. Look at the picture of the Quantas Jet. Any debris slung from an outboard engine in that 25 degree arc will hit the inboard engine pylon, wing and/or fuselage. The fan shroud is designed to contain one or two slung blades, but no part of the engine is designed to contain the entire disc or even a substantial part of a turbine disc. In the Quantas case the jagged metal sticking up from the top of the wing (below the "Au" in the logo) is the result of part of the turbine disc. What you cannot see is the more significant damage from approximately 1/3 of the intermediate pressure turbine disc detaching from the hub, and shearing through the engine case, cowling, lower wing surface, front spar, and the top surface of the wing.
    see the ATSB report AO-2010-089
    This site [https://www.atsb.gov.au/publications/investigation_reports/2010/aair/ao-2010-089/] has a link which is
    20MB PDF:https://www.atsb.gov.au/media/4173625/ao-2010-089_final.pdf
    Bottom line; to design a fan cowling substantial enough to contain such an failure would make the engine too heavy to be viable. As long as you can show the probability of such a failure which could involve catastrophic damage (fatalities and/or loss of airframe) is extremely improbably (<1e-09) AND is not attributable to a single event then the design is acceptable (FAR 23.1309)
    Differ

    ReplyDelete
  9. For whatever reason the Qantas scenario did not happen this time. Why can't you just accept that?

    ReplyDelete
  10. Everyone (here) is talking about the Trent 900, but the engine that failed was built by Engine Alliance. I can't find info to say if it was a GP7270 or a GP7277. (This is apparently a follow-on design from the GE-90)

    So while all the Rolls-Royce Trent stuff is interesting, it really doesn't apply to this incident.

    There are so many things that can go wrong with a high-bypass ducted turbofan, that it is amazing (to me) that they ever work. The exhaust temperature is higher than the melting point of some of the engine, and so cool air has to be introduced everywhere aft of combustion to keep the engines from melting. Occasionally they do melt.

    Because of that heat, they use ball bearings in the front of the engine and roller bearings in the rear, because they can expand... Very cool stuff happens in a high-bypass ducted fan.

    The fans have to be balanced to a ridiculous degree because of how fast they spin. (A little vibration...) Which can also cause problems if there is a defect or damage in a given blade. (One blade breaks, the fan becomes out-of-balance...)

    They are pretty incredible machines. Given that literally 1000s of planes are in the air everyday, the fact that there are so few problems is a testament to the power of engineering. (Somewhere between 14,000 and 18,000 flights per day are the numbers I can find)

    ReplyDelete
    Replies
    1. Thanks for the identification on the engine involved.

      The most amazing two things I ever learned about the turbines is that the blades are (were?) grown as a single crystal of metal as the only way to get the strength to weight ratio.

      I say "were" because a video I was watching on the Rolls Royce engines said the blades were hollow titanium, and I don't know if it's possible to grow them hollow. It showed a couple of workers each lifting a full sized turbine blade. They appeared to take no effort to lift. No cranes were involved and no apparent straining.

      As so often happens, the comments evolved in a different direction than what I was thinking which were (1) is this a normal amount of engine failure in a very large jet that hasn't been on the market very long and hasn't sold that many units and (2) could it be that the engine meets everything it's supposed to meet, but wasn't specified properly?

      Delete
    2. I note I said, The most amazing two things I ever learned and don't specifically say the second of the two is the next paragraph about the blades being hollow.

      More coffee!!

      Delete
  11. A History Channel clip on the subject of what goes into testing an engine. (This is the GE-90, but it is all I can find...)

    ReplyDelete
  12. All great comments. I spent 10 years as a Flight Training Instructor at Airbus, A320 and A330. My standard spiel to new trainees pointed out the fact that "you will probably never experience an engine failure" but we still train extensively for a failure at the most critical phase of flight: takeoff. US air carriers have an almost impossibly excellent safety record. It is a product of technology, superb maintenance, pilot training, air traffic control, &TC. But right at the top of the list is the incredibly reliable jet engine.

    ReplyDelete