7
$\begingroup$

I just finished reading this article and couldn't understand why in 3 minutes and 30 seconds even though there was no reliable speed data, pilot could not level the plane? Isn't the angle of attack reported by a gravity based tool like in off-road vehicles?

Summed up Question: When it's unclear what is happening with the aircraft, isn't it safer to put it in level position and medium speed? (Sorry for rudimentry language here)

$\endgroup$
  • 5
    $\begingroup$ Gravity doesn't work for this, as the load on an aircraft is not always 1G. $\endgroup$ – egid Jun 30 '15 at 23:56
  • 1
    $\begingroup$ @egid: I beg to differ - one of the primary flight instruments is the pilot's rear end. It is really easy to feel the difference in attitude between stall and level flight with your botty alone. I for myself have wondered just like Alexus how in hell these people could ignore such obvious indications. $\endgroup$ – Peter Kämpf Jul 1 '15 at 4:14
  • 5
    $\begingroup$ Peter, that's not always true. Accelerated stalls, stalls at very high speed (such as an airliner at cruise, etc) and the like will not have normal feedback. $\endgroup$ – egid Jul 1 '15 at 4:19
  • 2
    $\begingroup$ There was nothing stopping the crew from lowering the nose and flying out of the stall. In short, they got it all wrong. The angle of attack has no relationship to gravity. The angle of attack can be any angle relative to the Earth, and therefore to gravity. $\endgroup$ – Simon Jul 1 '15 at 4:55
  • 3
    $\begingroup$ Furthermore, they actually had reliable speed data the entire time that the aircraft was stalled. They just didn't believe it because it was such an unrealisticly low value and the stall warning had stopped indicating a stall because it was so low. There are significant human factors and design issues that led up to this accident. $\endgroup$ – Lnafziger Jul 1 '15 at 5:09
10
$\begingroup$

I just finished reading this article and couldn't understand why in 3 minutes and 30 seconds even though there was no reliable speed data, pilot could not level the plane?

If you look at this video by the French accident investigation bureau (BEA) you can see what happened.

The problem started when the speed indicators failed temporarily because of icing of the pitot tubes. The aircraft was flying through the intertropical convergence zone (ITCS) where severe thunderstorm activity can build up to high altitudes.

Instead of maintaining attitude and thrust setting, which is a normal procedure when speed indication is lost, the pilot flying (PF) brought the nose up. The aircraft started to climb but at the same time lost a lot of speed (trading kinetic energy for potential energy). Eventually the airspeed dropped so far that the aircraft stalled. By the time the aircraft entered the stall, the pitot tubes where de-iced and all speed indicators were working again. From the moment the aircraft stalled it started to sink. The PF tried to maintain a nose-up attitude, eventually using full aft stick. We can only guess to the reason why; it is likely that he was disoriented and had no confidence in the instruments after the failure. Yet it goes against basic flying instincts to pull the nose up in a stall.

The aircraft is naturally stable and will automatically drop its nose in case of a stall so it took some effort by the PF to keep the nose up. The horizontal tailplane then stalled as well which removed a lot of the nose-down tendency. The result was that the aircraft was in a nose-up attitude for the most part of the descent.

The angle of attack remained therefore too high and the aerodynamic stall persisted all the way down. If the crew would have brought the nose down they would have increased the speed, lowered the angle of attack and recovered from the stall. Once recovered from stall they could then level off the plane.

It may have contributed to the confusion of the crew that the airspeed was below 60 knots during portions of the descent. As they accelerated through 60 knots, the stall warning re-appeared.

The stall indicator systems makes use of alpha vanes which need some airflow over them to work. By inhibiting the stall warning below 60 knots the system is prevent from giving nuisance warnings on the ground. Such a speed is not normally encountered during flight, even not during a stall.

Another contribution to the accident may be that the pilot-not-flying (PNF) cannot see what stick inputs are given by the pilot flying. If it would have been in a different aircraft with coupled yokes, the PNF may have realized what was going on.

Isn't the angle of attack reported by a gravity based tool like in off-road vehicles?

The angle of attack is the angle between the chord of the wing and the incoming airflow. It is not related to gravity.

Summed up Question: When it's unclear what is happening with the aircraft, isn't it safer to put it in level position and medium speed? (Sorry for rudimentry language here)

Initially only the speed indicator failed. If the crew would have remained at the same attitude and power setting the aircraft would not have stalled. That is the standard procedure in such a case ("thrust and attitude") That may not have been the proper reaction in this case. The aircraft was encountering severe turbulence and the fastest way out may have been to climb. But after the stall warning sounded all speed indicators were back to normal and the aircraft was flyable.

$\endgroup$
  • 5
    $\begingroup$ Re conflicting stick inputs: The CVR transcript has six instances of "dual input" announcements. Also Robert earlier warned Bonin numerous times that he was climbing when he'd been told (and agreed) to descend. After Bonin said he had no control anymore and Robert said "controls to the left", Robert evidently did not use the stick priority button on his stick to disable Bonin's inputs when the aircraft warned him six times of dual inputs. CRM failure. $\endgroup$ – RedGrittyBrick Jul 1 '15 at 9:09
3
$\begingroup$

Something I've seen written but not discussed on many of these internet discussions, is the how the stall warning sound worked.

Apparently the stall warning sounded, but stopped again as the plane went to higher angles of attack. (This might be a deliberate behaviour, to prevent false warning sounds when operating on the ground in wind.)

As a pilot lowered the angle of attack, suddenly the stall sound started again. One of the pilots may have heard this and instinctively pulled up again to clear the sound. The angle was kept (very) high for the rest of the flight...

(I'd say the pilots had at least some reason for genuine confusion in this matter.)

$\endgroup$
  • 3
    $\begingroup$ Pulling up when you hear a stall warning is the opposite of what you should do, there's no reason the stick should have been pulled back again. $\endgroup$ – GdD Jul 1 '15 at 7:54
  • 1
    $\begingroup$ ...we all agree, sitting at home in daylight, but in their case, pulling back the stick cleared the stall warning. The warning behaviour probably contributed to the confusion of the one pilot who was pulling back. (There's more on the released transcript.) $\endgroup$ – Andy Jul 1 '15 at 8:05
  • 4
    $\begingroup$ I've read the transcript, my personal opinion is that there was a lack of fundamental flying skills which comes from operating a computer with wings. The urge to make an auditory warning go away is a product of that. Perhaps flying time without automation should be mandatory to keep pilots familiar with their roots. $\endgroup$ – GdD Jul 1 '15 at 8:46
  • 3
    $\begingroup$ @gdd It's easy to look back and say that, but keep in mind that they stalled in IMC conditions, over the water, at night, with what they felt were unreliable instruments. It was a terrible combination and the deck was definitely stacked against them. Yes, I agree that if they had understood what was happening that it would have been very easy for even many student pilots to fix. Instrument failures, that in real life can happen much differently than in training, can be hard to figure out, especially in the limited time available here. They made mistakes for sure, but not as easy to see. $\endgroup$ – Lnafziger Oct 31 '15 at 15:16

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.