Monthly Archives: July 2012

Graphic horror

AFR447 pitch order and sidestick deflection, from Bureau d’Enquetes et d’Analyses pour la securite de l’aviation civile

I’m still working my way through the final BEA report into the crash of AFR447. There’s certainly a lot to digest, particularly if you’re got any interest in human factors, and the relationship between automation and pilot training. But the thing that’s struck me so far is the image above, which came as part of the factual summary of the flight.

The image is a depiction of, among other things, the flight director orders to the pilots (where available) and the inputs given to the sidestick by the pilot flying the aircraft. Over the space of 90 seconds, the PF commanded no fewer than 27 significant pitch changes — the only thing I can think of is watching him push the sidestick back and forth without really understanding what was happening (and, given the CVR transcripts, that’s pretty much exactly what happened); perhaps unfairly, it brings to mind a driver who, skidding out of control on an icy road, repeatedly cranks the steering wheel back and forth.

I’m not qualified to comment on what those kinds of rapid pitch changes would do to an A330 in either flight law, or whether that kind of thing is normal.Based on my own experience, I’m guessing that rapidly switching between 1/2 nose up and 1/2 nose down, over the space of a couple of seconds, probably isn’t normal or something they teach you in Airbus Flying School. Having said that, BEA wonders whether or not the flight director’s commands — really, suggestions to the pilots — as represented by the green bands on the graph, might have lead the pilots further astray and been a contributing cause of the accident.

Learmount has more about this, and his point is well-taken:

If the flight management system computers have recognised their limitations and handed back control to the pilots, what are the flight director bars doing pretending they know any better? …

The flight director cross-hairs on the primary flight display are an aircraft-aiming tool that is increasingly irrelevant in an era when pilots, according to airline standard operating procedures, hardly ever touch the controls.

Yet they are a compulsively attractive tool. If they are there, you will follow them, especially if you are flying manually on the very rare occasion when you find yourself having to go manual in IMC. If you are feeling under-confident and out of practice, and are flying at night with no natural horizon, you will find yourself clutching at those crossed straws, even if they are mis-directing you.

We have no idea what the FD was telling the pilots — BEA had to rebuild the FD data based on an assumption about what the A330 would have been thinking, given the same air data inputs — and so we have no idea whether the pilot flying was just manipulating the sidestick back and forth frantically, or whether he was trying to follow a flight path laid out for him by the airplane, even as that path was essentially the exact opposite of what was needed under the circumstances, and even as the airplane itself was literally screaming “I’M STALLING!!” at him.