Hand fly using raw data. Cigar in mouth and John Wayne hat on. It is mandatory to remind the white faced FO on the way down that this is what you signed up for when you decided to fly and 38,000 hours later you still need to respect the airplane and the process and not take any shortcuts. You always fly the plane and never let it fly you...
Most accurate and lowest risk method, after all the plane will do exactly what the pilot commands.
If you believe that, I have swamp land in Florida and a bridge in Brooklyn that you might be interested in.
What amuses me in all these "perils of automatics" diatribes is that the posters never actually get around to any real quantification or analysis of the relative risks, nor do they reference the substantially lower accident rate associated with today's operations versus those of yesteryear. Automatics do bring with them their own risks, lowered manual proficiency and automation complacency being two of them. Those risks are, however, several degrees of magnitude lower in terms of their prevelance as causal factors than the risks that automation was intended to address.
With respect to this accident, up until Sitting Bull's recent post there has been nothing much to really go on but suppositions that were unanchored to any established facts,. And though some have done valiant work trying to find facts, the contributions of the "grumbling grandpas" and the "low skill libyan" contingent have really dragged the thread down. There may ultimately be some truth to both assertions, but so far there are no known facts that suggest either, and it just gets tiresome reading the same old dogs banging the same old drums again and again.
If we accept the following as reported by Sitting Bull:
-AAW771 was cleared for and executing the NDB 09 approach
-they did not report any problems nor did they request any medical assistance
-they announced and initiated a go-around at low level (tower radar and Alitalia TCAS)
-according to the official statement of the Alitalia crew that witnessed the crash, AAW771 descended out of the low clouds nose down, wings level and in one piece
-after violently hitting nose first it disintegrated with the tail separating and tumbling over
Then this, along with some previously suggested information regarding the crew's experience on type produces a plausible hypothesis that was suggested earlier on: somatogravic illusion.
The A330 (like many other heavy twins) has very substantial excess power available when operating on both engines at typical landing weights. When the pilot selects TOGA all of that power is brought to bear, causing both a pitch up moment and rapidly accelerating airspeed. Managing the change in attitude and airspeed, especially at lighter weights is more of a challenge (and much less practiced) than doing the same on one engine.
The significant acceleration involved can create a very convincing sensation of a more rapid than desired pitch up that can be quite disturbing for a pilot unprepared for it. This would be especially the case for a pilot with limited experience on the A330 but lengthy experience with 4 engined or lower powered twin engined aircraft.
Per the Alitalia crew's observations (TCAS) it would appear that a normal go-around (the crew called out that they were going around as opposed to a terrain avoidance manouever case which wouldn't likely have been called out) had been positively commenced and that the aircraft had begun to climb only to be observed descending with a substantial nose down attitude moments later.
The simplest explanation for how the aircraft got from A (climbing, hence nose up) to B (descending with nose down) is that either the pilot or the autopilot introduced a very large scale nose down pitch input shortly after commencing the go-around. It is not impossible that the source of the input was the autopilot, but it is much more likely that the source of the input will be found to be a pilot suffering from and responding to a somatogravic illusion during the go-around. This would also explain the degree of destruction of the airframe which is consistent with a high G impact (nose down attitude + TOGA = high vertical rate).
If this should all prove to be the case, the sad irony is that some of the old dogs and grumbling grandpas will have got it entirely wrong. The airplane may well have crashed, not because the crew relied on the automation too much, but rather because at the critical moment, they relied on the automation too little.
The somatogravic illusion has been around as long as there have been airplanes in the sky and its claimed victims from both old fashioned steam driven aircraft and now new-fangled electric jets. The problem is that no matter how big you make the primary flight display (ADI), how capable the autopilot, or even how often you hand fly your approaches all the way from TOD, you can't prevent the pilot from experiencing the illusion, you can only train him in how to recognize it and how to respond to it ... including by using an autopilot that does not experience the illusion in the first place.
ELAC