Submitted by Codehead on May 25, 2011.
How do you get into an unrecoverable stall, when you have 30,000 feet to recover in? Evidence shows that the Airbus did a "belly flop" onto the ocean.
How do you get into an unrecoverable stall, when you have 30,000 feet to recover in? Evidence shows that the Airbus did a "belly flop" onto the ocean.
I'd known about deep stall going all the way back to the BAC 1-11 test flights, but didn't understand how it applied to non-T-tail aircraft. True deep stall involves getting the horizontal stabilizer in or below the disturbed air from the wing. The BAC 1-11 crew simply could not get the elevators to exert enough force to force the nose down.
This website http://www.rbogash.com/Safety/deep_stall.html
has a great description of a pseudo-deep stall of a 727 caused not because the tail couldn't bring the nose down, but because the crew misunderstood what was going on. Like 447 they lost airspeed indications. Then, for some reason, they believed they were in danger of excess Mach, and they interpreted stall for high speed buffet. AOA was way too high, rate of descent was way too high, and yet they pulled back on the stick all the way to the ground. Ugh!
How else could a standard tail jetliner not drop it's nose for 30,000 feet?
It is of course much easier to blame the pilot and co-pilots who cannot defend themselves than to consider technical malfunction on Airbus and its suppliers and Air France. As for the factual report to be issued tomorrow by BEA, we can only expect "selected" information but no transcripts or recordings that will bring any new information as for the causes of this terrible accident. Just one month prior to the Paris Air Show what can we expect?
I'm not a pilot (just a retired aero engineer) so I have a question for pilots: Since flight simulators tilt the cockpit back to simulate acceleration, is it reasonable that an big AOA (25 degrees?) could be interpreted by a panicked pilot into thinking he was accelerating? At night, with no visual clues, with maybe too much going on to read all the instruments? And maybe thinking the plane was slightly nose down and picking up a lot of speed? Just thinking. And if so, might that pilot get obsessed about not going trans-sonic?
Pilots fly exactly as they are trained. For Airbus to try to blame the pilots is disingenuous. Airbus dictates the specifications of the simulators, the training syllabus, and writes the procedures.'
When an Airbus aircraft does not behave as Airbus predicts, the pilots are "on their own". They revert to basic airmanship honed over years and years. These were no neophytes...they had thousands of hours and had obviously gone beyond basic stall training. They did the best they could faced with a situation that Airbus never envisioned.
Airbus has always tried to design pilots out of its cockpits...the Toulouse accident was vintage Airbus logic. Once again, this philosophy has bitten them in the arse. Now they've got some s'plainin to do and they are yelling, "PILOT ERROR".
Yeah, right.
A deep stall behaves quite differently than a typical stall we all experience in primary and recurrent training.
A deep stall is normally an unrecoverable stall or may require some rather unorthodox maneuvering for recovery. It happens as has already been explained, when the horizontal stab enters the turbulence from the wing and the elevator can no longer exert enough force to bring the nose down. Typically the wing and tail are both stalled, and the wash or turbulence from the stalled wing prevents the tail from doing much of anything.
The tail normally exerts a negative force which keeps the nose up. IE It pushes down behind the CG. If the horizontal stabilizer (stab) stalls the nose drops. There is a balancing act between lift from the wing and lift from the tail. In a deep stall the nose is *usually* quite high and the horizontal stab exerts no lift, either up or down. In instrument conditions, or at night with no outside reference and with instruments lying to you, your senses (and the computer) are likely to interpret this as an acceleration.
Indications are that the plane impacted, or pancaked into the water. The only typical way this can happen is in a deep stall or high angle of attack. However if the plane has flying speed (under control) it will still have a substantial forward speed even with a high angle of attack.
This phenomena (nose high = acceleration) is how simulators can *simulate* acceleration by raising the nose. Conversely acceleration with no outside reference is interpreted as a pitch change. Like Vertigo, no pilot is immune to this. Add to that your instruments are lying to you and it's all over except for the shouting.
Another point to remember is that the larger the aircraft the more momentum it has. Even many light twins may take many thousands of feet to recover when all is going right.
First, stalling a jet is nothing like stalling your local 172. A large transport aircraft is an aircraft no one should ever stall. This is why they are equipped with stick shakers and pusher, as required by their type certificates, to ensure they are never stalled. An Airbus 330 is also a fly by wire aircraft. A feature of this aircraft is computer controlled control surfaces. The computer code governing the control surfaces is designed to never allow the aircraft to enter a stall or even an unusual attitude. Thus, for the aircraft to stall it necessitates that there must be a technical failure. The pilots failed as well if they allowed the aircraft to stall. However, if the technical failure fooled the aircraft to the extent that it would not accept the pilot inputs, there would be nothing they could do. That scenario is theoretically impossible though. If the computer receives data that is contradictory to flight regime limits, i.e., the aircraft accelerates at a high pitch angle with the power at idle, hence not possible, the computers go to a primitive mode in which the pilot inputs control the surfaces directly. At that point the computers will no longer interfere. Should an aircraft of this type already be in a deep stall by the time the computers give up control or by the time the pilots recognize what is happening, it may well be too late.