What Really Happened Aboard Air France 447?

[Dane_in_LA]

Originally Posted by eskercurve

After reading the Wikipedia article, I totally believe that at least one airspeed indicator was false and that caused the pilot to disconnect the autopilot.

If I understand the BEA preliminary report correct, the PF was doing a course change to avoid the worst of the weather.

Lack of outside reference horizon and inability to determine rates due to again lack of trusting the instruments probably led the pilot in command to repeatedly stall the airplane.

Looks like it. An A330 won't let the pilots stall it under normal law, but due to the iced-over pitot tube, the plane was in alternate law.

So the Airbus design isn't at fault per se here, especially if the pilot was able to stall the airplane. That tells me that the alternate law was engaged and flight envelope limiting was disabled.

Looks like it. The CVR recorded a snippet of conversation about it, even.

What Airbus could have done is address the issue of not doing good enough testing on the pitot probes. Those probes should be heated (I have seen them many times before and after a flight, those things are HOT! you can see the heat flow off them on a summer day).

They are heated, but apparently weather conditions were worse than the designers had tested for. (The pitot tubes were over-engineered when compared to FAA/EASA specs.)

All signs point to bad flight planning and bad decisions from the flight crew.

Again, looks like it.

The Airbus vs. Boeing vs. everyone else method of flight deck design and controllability philosophy doesn't really factor in here.

The only thing would be that the sidestick design w/o feedback made it impossible for the left-seat pilot to discern that the right-hand pilot was still providing nose-up input, even when he said he'd handed the aircraft over. Of course, people tend not to design aircraft around the idea of people in the cockpit doing the exact opposite of what they should be doing.

I would know that if there was an airspeed disagreement and the backup systems couldn't figure out who was right, I'd put the engines to max and trust my inertial reference system and climb. That cannot be fooled with by the wind and rain because those are (if they're modern) gyros which give actual attitude and rate information. Or I'd keep the autopilot on and climb. If the Airbus autopilot and inertial reference units are worth its weight in salt, it would maintain a reasonable flight path angle while climbing and using the acceleration data from the inertial reference unit.

FWIW, the pilots applied power (TO/GA). It just went to waste because they kept the damn thing in a stall. If they'd nudged the nose down, they would have regained control.

[eskercurve]

Originally Posted by Dane_in_LA

The only thing would be that the sidestick design w/o feedback made it impossible for the left-seat pilot to discern that the right-hand pilot was still providing nose-up input, even when he said he'd handed the aircraft over. Of course, people tend not to design aircraft around the idea of people in the cockpit doing the exact opposite of what they should be doing.

Yeah. That's an ongoing debate about which is better: a sidestick versus a traditional control yoke. Personally, I like the control yoke design better because it allows (artificially) augmented feel in both lateral and pitch commands. Having a sidestick, not so much. Having a sidestick pretty much forces envelope limiting for the very reason why this crash happened. Under normal VFR conditions they could easily feel the G's and refer to the horizon and cloud cover for orientation, but with this system in heavy clouds, yoikes!

Disclaimer: I work for Airbus' competition. They make OK airplanes. Given what I know of the features of both aircraft I would say both are pretty safe all things considered, but there are just some design decisions I wouldn't make so I prefer not to fly on them. Boeing planes have more parts in them so there's more potential for stuff to break, but Airbus planes just sidestep these issues in favor of other designs with their own quirks and in some cases dangerous situations. It's that 1 in a billion (literally) situation where I prefer to be in a Boeing airplane.

[DowntownHarrisburg]

What strikes me as odd is the co-pilot's first response. Pull back. This is more of a psychological question than anything else -- why did he pull back?

Was it some sort of physical reaction? As in, he tensed up and, given a mild rush of adrenaline, didn't realize that his arms were contracting and pulling the stick back? Anyone who's ever flown before can attest to the power of spatial disorientation. It sounds ridiculous to any non-pilot, but it's entirely possible -- and disturbingly easy -- to get your plane into a configuration where your physical senses tell you you're flying straight and level while you're actually climbing / diving / in a bank / inverted. That's why we have instruments, but this wouldn't be the first time that an otherwise skilled, rational pilot failed to maintain situational awareness.

Was it psychological? Maybe he was remembering Eastern 401 and trying to avoid a repeat. That's the flight that crashed into the Everglades because the crew was so busy attempting to fix a "landing gear problem" (ultimately an inconsequential burned-out light bulb) that they failed to notice their steady descent into terrain. I can appreciate, if not argue with, the thought: "Uh oh. Something unknown happened. I'm going to keep us from crashing into the ocean while my partner fixes things." Again, this wouldn't be the first time that an otherwise skilled pilot made a poor decision.

Situational awareness is one of the very first things you learn in flight training. It gets drilled into your head over and over and over and over and over and over and over again. When things aren't going the way you expect, it's time to step back and re-evaluate everything. Specifically, a failure to climb when the stick is pulled back is an oft-repeated scenario describing an impending stall. A loss of altitude with the stick back tells you something about your situation is dangerously wrong. It might be structural failure of the craft or its control systems, but it's more likely that you've done something wrong. Level out (or nose down if things are dire enough), manage power, regain lift, and restore your preferred flight level.

These pilots -- or at least the one pulling the stick back -- failed on those counts repeatedly.

I also have to wonder about the A330's separation of sticks. My flight experience is limited to Cessnas, which all have mechanically-linked yokes. Had the A330 had something similar -- even a set of servos in each stick keeping them artificially synchronized -- it would have forced the second copilot to realize what was happening.

[Dane_in_LA]

Originally Posted by DowntownHarrisburg

Was it psychological? Maybe he was remembering Eastern 401 and trying to avoid a repeat. That's the flight that crashed into the Everglades because the crew was so busy attempting to fix a "landing gear problem" (ultimately an inconsequential burned-out light bulb) that they failed to notice their steady descent into terrain. I can appreciate, if not argue with, the thought: "Uh oh. Something unknown happened. I'm going to keep us from crashing into the ocean while my partner fixes things." Again, this wouldn't be the first time that an otherwise skilled pilot made a poor decision.

Some have speculated that AF pilots focus on training low-altitude impeding-stall scenarios (that's what can bite you the hardest, so you train for it), and he reacted according to that. Spin up the engines and gain altitude. Which is logical on approach, not at 35,000 feet.

I also have to wonder about the A330's separation of sticks. My flight experience is limited to Cessnas, which all have mechanically-linked yokes. Had the A330 had something similar -- even a set of servos in each stick keeping them artificially synchronized -- it would have forced the second copilot to realize what was happening.

FWIW, the only Airbus pilot I know love the entire setup, including the sidesticks. He gets his stick-and-rudder time in his Beechcraft, he's quite happy to fly what he calls "an old man's plane" for a living.

But it does take cockpit management. The left-hand pilot took over the plane - or thought he did. The right-hand pilot kept pulling up after that. It's tough to design an aircraft around the idea of a supposedly skilled operator consistently doing the wrong thing.