Thursday, October 29, 2015

FLT 800, et al

      WHAT THIS GUMMINT DOES
        

As previously mentioned, I find Smithsonian's "Air Disasters" series compelling. Very.  I just watched VANISHED.  An Airbus A330, traveling from Rio de Janeiro to Paris, vanished somewhere over the Atlantic. After two years of scouring the ocean floor they finally recover the two black boxes and discover the reason. 

Ice crystals temporarily clogged the device that measures air speed.  A common occurrence that will self correct in about a minute if level flight is maintained.  However, the Auto Pilot is automatically disengaged with an audible warning,  What happened was that the experienced head plilot had taken a sleep brake, leaving two "2nd officers" in charge.  When the alarm sounded, one pulled back on the yoke to gain altitude which, in the rarefied air at 35,000 feet put the plane into a stall.  Duh. The pilots, instead of nosing down to gain airspeed and lift,  kept it nose high while it fell 13,000 ft/min, kerplunk.  The reason?   Air France pilots had grown so used to the Auto Pilot that they had not been trained on manual flight in this event.  I am not making this up. 

Another reason I so love this series is that every episode is solved using using data.  Blame is put without fear of embarrassing any person, company, or manufacturer. And guess what?  So far not a single crash investigation has been performed by the FBI, with the final report produced by the CIA! 

So I checked.  Sigh.  TWA 800, has so far been ignored.

11 comments:

Steve in Greensboro said...

Known to frequent fliers as "Air Chance".

Anonymous said...

There's a video on YouTube that features Capt 'Sully' of the Hudson Landing fame, explaining why this crash probably would NOT have happened if they had been flying a Boeing aircraft. Boeing uses YOKES for the control of the aircraft, while Airbus uses Fighter-style SIDESTICKS. While the side sticks may have a 'cool' factor about them THEY ARE NOT PHYSICALLY CONNECTED! That's how the junior-most pilot who had control of the plane was able to- for several minutes- pull BACK on the stick, the absolutely incorrect procedure, while the aircraft continued to plummet without attracting the attention of the other, more experienced pilot.

As Sully points out in a Boeing simulator, when one control yoke is moved by one pilot, the other pilot gets immediate feedback because his yoke also moves in an identical manner, thus contributing to overall situational awareness. The more experienced pilot 'probably' would have detected that the junior pilots actions were incorrect and would have taken control of the situation. Once again, a good reason to stick with an American design.

Anonymous said...

Found it! Here's a YouTube link to Sullenbergers controls explanation...

https://www.youtube.com/watch?v=kERSSRJant0

Anonymous said...

But lowering the nose at the point of approaching a stall is about lesson 3 in the Cessena 150. It boggles the mind that someone with an ATP could ever be so out of sorts. Mentally disconnected from the aircraft is more likely. Somewhere in Youtube land is a video of a Senior AA Capt. scorching a room full of pilots for being "systems operators" instead of aviators. He states that several crashes were due to the pilot not being aware of the aircrafts trim state. In other words "George" has been flying for hours and the fuel burn has shifted the CG of the aircraft dramatically. The autopilot has been compensating for this out of trim situation. Then first officer Pedro turns off the autopilot without having his hands securely on the controls and the plane pitches violently, unopposed, into an unrecoverable situation. Can't seem to find the video. -Anymouse

rickn8or said...
This comment has been removed by the author.
rickn8or said...

And here's TWA-800:
https://www.youtube.com/watch?v=8cjbyJ8ELxw

(Spoiler alert: center wing box fuel vapors ignited by arcing fuel probe.)

Rip Tide said...

Takeoffs are optional, but landings are mandatory.

Wabano said...

At the first Airbus 320 crash in Paris, the pilots DID add power and pulled up, but the auto pilot, that could not be disconnected had decided to land in the trees...I have seen the video from helicopters showing Airbus employees stealing the black(red) boxes for trafficking them, then the crew were both arrested and jailed in an insane asylum...You have indeed to be insane to fly on Airbusses, after dozens of similar crashes...
Airbus pilots do not know how to fly a real airplane, so every six months, they are made to fly a simulator that will monkey how a real plane do...at $5,000 an hour, it would be cheaper to rent a cessna...

Anonymous said...

I've read that any modern commercial pilot, flying either Boeing or Airbus, only has hands-on-the-stick for about 5 minutes, regardless of the length of the flight. About 2 minutes for takeoff and three for landing. Everything in-between is flown by 'George', the auto-pilot. The main reason is for fuel economy, as the climb to altitude, transit and descent are all programmed to squeeze every penny out of a pound of fuel.

Airbus was the first to embrace this highly regimented flight automation, and Boeing has had to adopt it as well, for competitive reasons.

The age of seat of the pants flying in the commercial realm is dead, and with it, the skill set of the pilots.

TheOldMan said...

Re: AF crash
I don't quite understand why the pilots did not feel the aircraft sinking. It was not gradual, they must have felt the descent and yet they did not react. That is baffling to me.

Anonymous said...

In their defense, the crew had just been thru a high level storm that is often seen in the Intertropical Convergence Zone (ITCZ). This is where the confluence of warm and cold air meet where the northern and southern trade wind patterns collide near the equator. You've seen this on a smaller scale when tornadoes are produced. From the satellite data forensics made after this flight, the storm was hundreds of miles wide, and reached all the way to 50K feet or more. Just prior to this, the junior most pilot, who was flying the plane, asked 'What is that smell?'. The more experienced pilot responded that it was ozone, produced by the rapid lighting flashes within the storm. You can imagine that, in a darkened flight deck, it must have been like sitting in front of a strobe light. The supercooled air and moisture produced by the storm at altitude is what froze the pitot tubes that supply velocity data to the autopilot.

At the very time when the pilots needed the autopilot it abandoned them, and dumped the whole mess into their laps. This is one area that is being looked at in the aftermath of AF447, where those who design the automated systems ask themselves what more can be done to alleviate the issue. For example, first warning the pilots that the pitot data is inaccurate, but continuing to use the data from moments before, sort of a 'pause' feature. The autopilot would still fly the plane with inputs from GPS and Angle of Attack sensors to ASSUME the airspeed was within limits. GPS does not give you true airspeed, only the pitot sensors do, but it's better than nothing FOR SHORT PERIODS of TIME.

As for the junior most pilot pulling back on the stick when he should't have, it has been noted throughout aviation history that in high-stress situations pilots sometimes perform the wrong actions. This will only be solved by MORE training and actual hand-on-stick time. But it looks like the industry, due to market forces, is moving in the opposite direction.

Post a Comment

Just type your name and post as anonymous if you don't have a Blogger profile.