Special Pages

Saturday, April 21, 2018

About that Southwest Airlines 737 Engine Failure

Engine failure... sounds so sterile; so mundane doesn't it?  We can't call it an explosion because there were no chemicals exploding; no fuel going off.  It sounds like a joke to call it a "spontaneous engine disassembly", although that's fairly accurate.  Everyone knows the left engine on this Southwest Boeing 737 lost a turbine blade, leading to a very bad outcome: metal penetrating into the pressure vessel and killing a passenger.  I'm sure everyone has seen the story.

While I never worked on engines, I worked in the civil aviation industry for 20 years and have some knowledge about how the industry and FAA work.  Because of that I have some thoughts I want to share with you.

Let's start here: there's a reason that Jennifer Riordan, the woman who was killed, was the first U.S. passenger airline fatality since 2009: the relentless drive of tens of thousands of engineers of all kinds driving to always determine the root cause (brief overview) of all problems and to design systems so that a single failure, like losing a turbine blade, doesn't take down an airliner.  The engine housing should not have allowed the debris that punctured the fuselage to escape: that means there were two failures here.  The second failure, and arguably the one that caused the fatality, was the failure of the engine housing to contain the flying debris.

All jet engines on commercial airlines get qualified this way.  For example, here's a video of one of the huge engines on an Airbus double decker A380 having a turbine blade blown off with an explosive as part of its qualifications test.  Engines are designed to survive this; to fail gracefully.  Additionally, all two engine aircraft like the Southwest 737, are certified to be able to fly on one engine, including be able to climb out from an airport should the engine fail immediately after takeoff.

There's an undercurrent among the news talking heads that they're scared this is going to start happening widely.  I'd say that's not likely.  The Boeing 737 is one of the most common aircraft in service, with production lines rolling out about two new aircraft every day.  I don't know what percentage use the CFM56-7B (GE/Safran) engines that are being grounded for inspection, but it's not a big population of engines they're inspecting.  Reuters put some numbers in the story.
Ultrasonic inspections on fan blades that have been used in more than 30,000 cycles, or in service for about 20 years, will be required in the next 20 days, the agencies said on Friday. A cycle includes one take-off and landing.

That order will affect about 680 engines globally, including about 350 in the United States, the FAA said. The engine that blew apart on Tuesday’s Southwest flight would have been affected, since the company said it had 40,000 cycles.
680 engines in the world, somewhere between 680 and 340 aircraft, when at any given moment there are 5 to 10,000 aircraft in the air.  The chances of even being on one of those aircraft are vanishingly small. 

Everything that gets made can fail.  System engineers compensate for that by designing systems so that it takes more than a single point failure to take out the system.  If the two failures are truly independent of each other, the probability of both failing is very small. 


Being improbable doesn't mean it can't happen; after all, someone wins the Powerball lotto regularly.  A failure analysis is going on here, and it's a safe bet there will be a change to something.  Perhaps just a change to inspection frequency or methods for the turbine blades or a change to allowed hours of flight; perhaps design changes to the forward part of the CFM56 engines, perhaps something else.  This failure will become less probable.    


11 comments:

  1. The turbine section looks intact. Fan section lost a blade for certain, not certain if the compressor section lost containment.

    ReplyDelete
    Replies
    1. Hi "Aggy!!!,"
      "10-4!!" "BUT," we have to put up with the "Arm Chair Quarterbacks" that have never even "Flown" in an Airplane!!! As an old "Aerospace Type" I earned my wings working for the likes of "Leroy's Iron Works," (Grumman Aerospace F-14IRAN project) Douglas Aircraft (Don't you dare call it McDonnell-Douglas to the "Old Farts!!") ACES II EJECTION SEAT project, WEBER AIR in beautiful "UP-TOWN" Burbank ,,,also ACES II, SCURVYIRVIN AERO (I Packed The FIRST SPACE SHUTTLE DRAG CHUTE ON ENDEAVOUR!!!!!!!!!) Go figure.... ???? and yes, the root cause will be found out!! "Check ad double check, every time you put your QCstamp on a "buy off" you better lick it and kiss it that you are right.... 'Being a "PARACHUTE RIGGER" ... If I won't jump it, .. well.. GO FIGURE!!
      Skybill,
      FAA Private Pilot (SEL), MASTER PARACHUTE RIGGER (Chest, Back and Seat) and USPA LICENSED SKYDIVER, D-6009.....also SCR-2034, SCS-680 and "AIR TRASH" (In Good Standing!!!!!)-- want more info?? go to www.airtrash.com and "Browse!!"

      Delete
    2. Fan section lost a blade for certain, not certain if the compressor section lost containment.

      Yes, but. The nacelle is supposed to contain the debris from the fan section. There's several videos of engine qual tests where they blow out a turbine blade and the nacelle captures the debris with no trouble.

      every time you put your QCstamp on a "buy off" you better lick it and kiss it that you are right. That's for damned sure! Been there, done that, too, although not in aviation. Just process control hardware going in to a nuclear power plant to monitor everything, or into a plant making chlorine gas.


      Delete
    3. I was pointing out you said turbine lost a blade...

      Delete
  2. I watched the video and listened carefully to the math.
    It is just as well that a much younger me didn't know exactly what dragons were sleeping inside the Navy's main propulsion steam turbines.

    ReplyDelete
  3. https://www.youtube.com/watch?v=4a4Dw-IbsKg

    Start around 3:00 if you want the gist of the failure. It was not the BLADE which penetrated the aircraft. It was instead pieces of the engine inlet section. The blade failure itself was contained.

    Mr. Brown is a good source of information on civil aviation, and is also closely following the Oroville entertainment.

    ReplyDelete
    Replies
    1. Good video and excellent description of all the frantic motion going on in the cockpit.

      Delete
    2. By the way, from what he says, I'm nor sure if anything actually got INTO the cabin area. If the debris from the engine inlet fractured the exterior window pane to the point it failed catastrophically, the plastic inner pane would have blown out, since it is only designed to prevent damage to the exterior pane and NOT to be able to structurally contain the cabin pressure at flight level 320.

      Delete
  4. In one respect this is like school shootings. Statistically uncommon but dramatic and good fodder for headlines and cable news. So as a result of getting 100% of the news for a long time people are scared by it. The biggest risk in air travel is the drive to the airport. No mention of that. School bus accidents probably kill more school children than AR15's do but no mention of that. High school kids texting while driving probably kill many, many more kids than AR15s do and no mention of that either.

    A co-factor in this is that if this woman killed in the airplane accident were killed while driving to the airport all of the ambulance chasing lawyers would say "ho-hum". But this is the big one. This could put a single lawyers kids through college and buy him a new house too. This requires hype and news attention.

    No word from camera Hogg if he is going to demand we ban airplanes.

    ReplyDelete
  5. After retiring from a 23 year career in Aerospace as a Manufacturing Engineering Planner (ME Planning creates and maintains the build plan and Manufacturing Bills of Material) @ MDAC and Boeing I am still amazed that anything as complex as a modern airliner can function at a level so close to perfection.
    _revjen45

    ReplyDelete