AF66 CDG-LAX diverts - uncontained engine failure over Atlantic
Join Date: Dec 2006
Location: UK
Posts: 29
Likes: 0
Received 0 Likes
on
0 Posts
RE the comments about stopping the CVR in flight ..I don't think an SOP of pulling the CVR recorder in flight is a good thing at all. With the benefit of hindsight doing so might have been in the current AF investigation but what about a hypothetical incident which unlike this one doesn't end well, and unfortunately ends up with a crumpled heap of metal under the approach or off the overrun.
Let's say an engine goes pop as it did in this case, the issue is contained OK initially and a decision is made to pull the CVR.....since the CVR was stopped we now have a well preserved record from the cruise of a thump, somebody calling for a severe damage separation checklist, a bit of T-DODAR....and then nothing....no record of any subsequent analysis done by the crew, any reassessments during the descent, any descent/approach briefing, any conversation/comments made during the approach.
There certainly seems to value in increasing the CVR recording time, and in any event a lot will be gleaned from the flight data recorder....but pulling recorder CBs in flight?
Let's say an engine goes pop as it did in this case, the issue is contained OK initially and a decision is made to pull the CVR.....since the CVR was stopped we now have a well preserved record from the cruise of a thump, somebody calling for a severe damage separation checklist, a bit of T-DODAR....and then nothing....no record of any subsequent analysis done by the crew, any reassessments during the descent, any descent/approach briefing, any conversation/comments made during the approach.
There certainly seems to value in increasing the CVR recording time, and in any event a lot will be gleaned from the flight data recorder....but pulling recorder CBs in flight?
Join Date: Oct 2003
Location: n/a
Posts: 1,425
Likes: 0
Received 0 Likes
on
0 Posts
The CVR recording times compared to changing technology have been a joke for years. However, they are finally changing and 20 hour CVRs will become norm on new aircraft... but not until after 2020 (ish).
For US-registered aircraft, 14 CFR 121.359 has required 2-hour recording since 2012. Most Airbus (certainly A380) also have 2-hour recorders. QF32 had a 2-hour recorder but the recording of the actual incident was still overwritten because it took so long to suppress engine #1.
It's easy to say "look at my new toy MP3 recorder that can record 10 years of audio". Your MP3 recorder isn't required to be certified under stringent regulations (which are there for a purpose) including being able to survive crash, fire, submersion at high pressure, etc., and emerge with readable memory. What's more, MP3 is very compressed; I don't know, but I expect CVR recordings aren't compressed because compression could fail and make the recording unreadable. Uncompressed audio needs a lot more memory.
Having said that, the blame culture in many airlines, even among pilots, creates pressure to limit recording time and make recordings erasable. Even here on what's supposed to be a professional pilots' forum, the lynch mob was out for those CA pilots after their famous go-around at SFO almost before their engines had stopped. Fix that blame culture and you go a long way to making long-period CVRs acceptable.
It's easy to say "look at my new toy MP3 recorder that can record 10 years of audio". Your MP3 recorder isn't required to be certified under stringent regulations (which are there for a purpose) including being able to survive crash, fire, submersion at high pressure, etc., and emerge with readable memory. What's more, MP3 is very compressed; I don't know, but I expect CVR recordings aren't compressed because compression could fail and make the recording unreadable. Uncompressed audio needs a lot more memory.
Having said that, the blame culture in many airlines, even among pilots, creates pressure to limit recording time and make recordings erasable. Even here on what's supposed to be a professional pilots' forum, the lynch mob was out for those CA pilots after their famous go-around at SFO almost before their engines had stopped. Fix that blame culture and you go a long way to making long-period CVRs acceptable.
Last edited by OldLurker; 3rd Oct 2017 at 10:58. Reason: correction
Your MP3 recorder isn't required to be certified under stringent regulations (which are there for a purpose) including being able to survive crash, fire, submersion at high pressure, etc.,
However fundamentaly we increasingly live in a society where many seem to think the number one priority in any situation is to have a recording, or access of a recording, of an event. I assume the idea is this can then be used for a variety of purposes ( monetary, regulatory, increasing hits on social media outlets). Frankly I'm not keen on moving much beyond where we are now.......guarantee privacy, fix the blame culture, then it will be easier to make long period DVRs acceptable.
Join Date: Jul 2014
Location: Germany
Posts: 344
Likes: 0
Received 0 Likes
on
0 Posts
Join Date: Dec 2012
Location: France
Posts: 5
Likes: 0
Received 0 Likes
on
0 Posts
Pax2908's drawing shows exactly what was argued earlier. The LP compressor hub flange is fixed to the rear of the fan disc and is in fact only supported by the fan disc. When the fan separates from its shaft one would expect the LP compressor hub (together with its 5 stages of compressor blades) to be thorn out of the engine but instead the compressor hub mounting flange separated quite nicely, despite the many fastening bolts.
Join Date: Jun 2010
Location: The Netherlands
Posts: 18
Likes: 0
Received 0 Likes
on
0 Posts
Join Date: Mar 2015
Location: Coastal Georgia
Age: 71
Posts: 40
Likes: 0
Received 0 Likes
on
0 Posts
Fasteners gone from the LPC inner support structure/rings also.
Failure was not likely caused by a few unnoticed missing fasteners during assembly/inspection.
Photo link:
http://www.pprune.org/attachments/ru...lantic-one.jpg
......
Failure was not likely caused by a few unnoticed missing fasteners during assembly/inspection.
Photo link:
http://www.pprune.org/attachments/ru...lantic-one.jpg
......
Last edited by number0009; 3rd Oct 2017 at 14:15.
Join Date: May 2010
Location: Boston
Age: 73
Posts: 443
Likes: 0
Received 0 Likes
on
0 Posts
True and even with engines mistakes have been made. But engines are very critical parts so no margin for such errors, very strict procedures and inspections. Also there's a lot of electronics monitoring the performance of the engine, I would think that "half of the bolts missing" would cause a lot of vibrations and make all the alarms go of when they do a test run of the engine. But indeed never say never...
More details at:
https://en.wikipedia.org/wiki/NOAA-19
This type of "mishap" is sometimes caused by interrupted or resequenced work.
Join Date: Jun 2010
Location: The Netherlands
Posts: 18
Likes: 0
Received 0 Likes
on
0 Posts
Join Date: Sep 2015
Location: UK
Posts: 110
Likes: 0
Received 0 Likes
on
0 Posts
Depends on your definition of "a lot". An hour of uncompressed CD quality stereo audio is 527 MB. My current iPhone holds 64 GB of storage, that's equivalent 124 hours of uncompressed stereo CD quality audio. The memory requirements are trivial in this day and age. Solid state digital CVRs exist and are certified, all you need to do is stick more memory inside them.
Join Date: Aug 2003
Location: FR
Posts: 234
Likes: 0
Received 0 Likes
on
0 Posts
#9 in "your" photo I think one sees clearly two types of holes, there are 24 which seem "empty" (no fasteners there, I suppose these only are there to clear the 24 fasteners that belong only to the fan section, see also the photo showing the engine in the shop with only 6 fan blades on); and then there are 3*24 holes which ~all seem to show an end of the fastener still inside ... so indeed would seem as a clean break (perhaps as per design).
Fasteners gone from the LPC inner support structure/rings also.
Failure was not likely caused by a few unnoticed missing fasteners during assembly/inspection.
Photo link:
http://www.pprune.org/attachments/ru...lantic-one.jpg
......
Failure was not likely caused by a few unnoticed missing fasteners during assembly/inspection.
Photo link:
http://www.pprune.org/attachments/ru...lantic-one.jpg
......
Join Date: Mar 2015
Location: Coastal Georgia
Age: 71
Posts: 40
Likes: 0
Received 0 Likes
on
0 Posts
pax2908,
Thanks, see that every 4th hole is used for the fan/LPC attachment most appear darker as if a larger dia.
Main point of photo link was to show what appear to be the missing or sheared fasteners of the inner compressor rings which would have all holes filled when assembled.
......
Thanks, see that every 4th hole is used for the fan/LPC attachment most appear darker as if a larger dia.
Main point of photo link was to show what appear to be the missing or sheared fasteners of the inner compressor rings which would have all holes filled when assembled.
......
Join Date: Mar 2007
Location: In my head
Posts: 694
Likes: 0
Received 0 Likes
on
0 Posts
Big pity, I read earlier, that the forward facing camera doesn't cover the outboard engines. Had it been otherwise, it might have made an eminently watchable engineering training video for decades afterwards!
Earlier in the thread some postulated that the majority of the missing lump would likely have flown off safely forward and down.
I have my doubts that any of it took that route.
Looking at the cleaner breaks at the bottom half of the remains, and the rent upward pylon fixings and the leading edge dents, my guess is that the fan in the fancase let go first at the bottom and folded up and then ripped off top-last over the top of the wing - if you blinked you missed it no doubt!
Earlier in the thread some postulated that the majority of the missing lump would likely have flown off safely forward and down.
I have my doubts that any of it took that route.
Looking at the cleaner breaks at the bottom half of the remains, and the rent upward pylon fixings and the leading edge dents, my guess is that the fan in the fancase let go first at the bottom and folded up and then ripped off top-last over the top of the wing - if you blinked you missed it no doubt!
Join Date: Jun 2001
Location: Rockytop, Tennessee, USA
Posts: 5,898
Likes: 0
Received 1 Like
on
1 Post
Looking at the cleaner breaks at the bottom half of the remains, and the rent upward pylon fixings and the leading edge dents, my guess is that the fan in the fancase let go first at the bottom and folded up and then ripped off top-last over the top of the wing - if you blinked you missed it no doubt!
Well, I am expecting that the fan went whichever way it felt like, and possibly whole. The close fit in the front cowling structure will have seen pretty much instant wrenching of that whole assembly from the main pod when the fan moved laterally enough to chew into it on its way out. 
Thanks for the diagram pax2908.
Thanks for the diagram pax2908.