My apologies to G-man. I hadn't had a chance to read that link until now. Upon review of the CVR transcript, I am changing my opinion on what happened. Saying there was an over reliance on automation is too simplistic, and ignores the total breakdown of Crew Resource Management that happened. Indeed, the right seat FO appears to have completely panicked and in his panicked condition he continued to make irrational control inputs that doomed the flight.
Here is how I see it:
As with all accidents, there was a chain of events that led to the unfortunate demise of AF447. The first decision was on the part of the captain in deciding to plow through the ITCZ when all the other flights along the route that night were diverting 75 miles to the west in order to avoid the Meso-Scale Complex of thunderstorms. It is pointless to drive right into the ugly black heart of a big area of embedded Cb's when a diversion of less than 10 minutes will keep you in clear skies. Doing it at night is even more pointless, and I am left scratching my head wondering why the Captain thought he should subject his passengers and crew to an E Ticket ride. Rule #1 of heavy weather flying in RC's handbook: There's no substitute for the Mark I eyeball when it comes to thunderstorm avoidance.
While I do not advocate penetrating Thunderstorms at night, it can be done provided you keep to the fundamentals. What is telling however is that the transcript shows that the right seat FO, is clearly nervous about what they are flying into. The FO queried the Captain about the St Elmo's fire that was present in the cockpit, having never seen it before. Just before they encounter the first cells, the senior FO returns to the cockpit, and the Captain turns over the left seat to him and goes back to the cabin for his rest. Now, would I have done that? Probably not. I would have been way too uncomfortable leaving the cockpit less than two minutes before hitting heavy stuff. I'd have stayed in the left seat at least until my plane was out of the heavy weather, then I would have let the senior FO relieve me. Apparently the Captain felt the senior FO and junior FO had enough experience between them that they could handle any situation that may have come up.
A little later (2:07) the senior FO figures out that their onboard radar was NOT properly set up, and discovers they're about to hit some serious cells! Major whoops! Still not a killer though. Around this time the PM author explains that the junior FO (who is the Pilot Flying, or PF) sounds increasingly nervous. From what I'm reading, he's literally jumping at every new light, sound, or sensation. He's just about to lose his self control.
Now at 2:08 the senior FO asks the PF if he agrees they're in Manual mode. I would assume this to mean the autopilot is off, and the junior FO should be hand flying the airplane. One never tries to fly through a thunderstorm on autopilot; they're not built to handle the stresses and usually have several safety features that will kick them offline once the turbulence gets bad. The article makes no mention of the junior FO's reply, but the airplane was still on autopilot.
Once the pitots iced over, the autopilot did kick off. This is normal and should be expected in severe weather. I have had pitots ice over and watched in amazement as my airspeed indicators both went to zero. However, the normal, logical thing to do in such a situation is do not make any major changes in attitude. Just keep the wings level and wait for the pitot heat to melt the ice off. Unfortunately, the junior FO did the worst possible thing: He pulled full back on the sidestick. This is a panic reaction that has been seen in other accidents where the airplane did something a nervous pilot wasn't expecting. Just look into what happened at Buffalo with Colgan 3407 (another panic reaction - full back on the yoke).
The senior FO kept trying to talk the junior FO into lowering the nose. However, as soon as the FO made his panicked reaction (full up on elevator) the guy in the left seat should have said "I have the airplane!" Indeed, the senior FO should have called "My Plane" as soon as the autopilot kicked off and the airspeed tapes went to zero. Instead, he rang back into the cabin for the captain! Unbelievable!
Investigators and the author of this article both have pointed out that the Airbus' fly by wire sidestick controllers are not interconnected, so it would not have been apparent to the senior FO exactly what his right seater was doing with the controls. But he was the acting PIC at that time, and he forgot that in his haste to get the Captain back to the cockpit. This was a terrible lapse of CRM.
Because of the full up elevator command the airplane zoomed up nearly 3000 feet and then entered a Stall. It remained in a stalled condition due to both pilots pulling back on their sidestick controllers. The article pointed out that the stall warning system called out "Stall" no fewer than 75 times before they impacted the ocean, yet neither pilot recognized the callout! I'd like it if Busdriver would give us a technical description of the alternate law fly by wire control system. It is apparently everyone's opinion that these pilots thought they could not stall the airplane (true in normal law operation) but this assumption was invalid because they were in alternate law. I have read other posts by other Airbus drivers that alternate law was discussed in both initial and recurrent training in the USA at that time. By the way PM's article is written, the author apparently thinks Air France's training may not have covered this. I'm not sure about this.
So you have one FO who is in a full-blown panic, pulling back on his sidestick. The other FO seems to be so confused that even when he does call, 'My Airplane!' he also pulls full back on the stick! Both are making irrational control inputs - some would call it the definition of insanity - just sitting there doing the same thing over and over and expecting the airplane to react differently. Neither one is in Command of the plane... they're both waiting for the Captain to get to the cockpit and save their collective bacon. Also, despite the senior FO announcing control of the plane, the junior FO jumped back onto his sidestick and continued pulling back on it!
Once the captain finally did get to the cockpit and figured out what was happening, yes it was too late. They're stalled and descending in excess of 10,000 feet per minute. The captain finally realizes what's happening as they descend through 10,000 feet and says "don't climb." It is about this time that the junior FO finally announces 50 seconds before impact that he has been pulling back on the sidestick the entire time! He is so completely panicked that even after he's been told a second time to get off the controls he got back on his sidestick again right before impact.
Everyone will draw different conclusions to this accident, but my own are this:
1) The crew had an inadequate preflight weather briefing, and/or did not wonder why all other traffic that night was diverting.
2) The Captain relinquished his seat to an FO right before they entered heavy weather. This is a no-no in my book.
3) When a primary flight instrument failed, a very nervous junior FO panics and pulls the Airbus into a stall, and does not realize what he has done.
4) An equally rattled senior FO does not exercise his duty as Pilot In Command as he should. Apparently he did not realize the Pilot Flying was in panic mode. Perhaps because he too was waiting for the Captain to return to the cockpit?
Overall, I get more of a sense of complacency that is the result of all the whiz-bang avionics in our flying machines. Since nothing ever goes wrong, what's the point of drilling all this stuff in recurrent? Perhaps there is too much emphasis on systems training and not enough on basic airmanship. We sometimes forget that we are hurtling through the air many miles above the surface of the earth, travelling at great speed in an environment that will kill us if we are not vigilant. When something does go wrong, it seems to surprise even the most experienced crews but this panic attack is indicative of something worse. The breakdown of CRM is an indication of a systemic problem within Air France's training, and I hope they are looking very closely at their operations and training environment and make changes accordingly.
Thread Drift over.
RC