Zorin 75:

The flight engineer tried to tell the captain he had lost an engine. The captain couldn't figure it out. The autopilot put in more and more aileron until its limits were reached, at which point it disconnected and induced a wing over.

As you know part of the tail came off during the ensuing high speed dive/high G recovery. The fact it didn't end up in the ocean was a tribute to Boeing, not the clueless pilots.

Here's the dramatic version:




Perhaps this incident isn't that far off topic at all. It too starts out with a comparatively benign failure. The PF is fixated on one problem (speed), breaks his scan and misses the plane's attitude, which is starting to roll. The a/p tries to correct for this, and when the PF takes over manual control he's taken by surprise. At this point they're so disoriented that they rather believe all their ADIs to have failed than actually be doing an aileron roll...

Extraordinary recovery and landing by the Air China crew.

But am i correct i thinnking that the Air France deep stall would have been much harder still to recover, being in a nose up attitude?

# 1756 refers.

It might easily be true that a recovery from that deep stall was impossible...I spent the last eleven years of my career in command of a jet airliner that was known to be impossible to recover from a deep stall...A stall recovery was only achievable from the early stages of the stall.

predictorM9;

I suggest you have a look at post #1817 AF447 Thread No.3 and check out the two links at the bottom.

The aircraft appeared to have a tendency to roll to the right, though this shouldn't have been due to fuel imbalance as that is equalized automatically. There is nothing in the BEA update to suggest a reason for the asymmetry.

Even with the THS trimmed 13° NU, the elevators will have had enough authority to cause a ND trim at IAS < 180 KTS. When that was done the stall warning sounded and the sidestick inputs were reverted to NU. The flight law was ALT 2 and the autotrim can be overridden by constant sidestick inputs or the mechanical trim wheel.

Once the aircraft had fully stalled at FL380, the AoA value was deemed invalid as the mean CAS was < 60 KTS. So you can see that after stalling they found themselves on the wrong side of the stall warning and didn't realize it.

Essentially the aircraft maintained >15° NU attitude all the way down and as the air became denser the forward momentum was reduced, the AoA increased while the RoD remained near constant.

Thanks MM43, I understand better about the stability.
But still, in post 1393 it says that the elevators are also fully stalled, so can they still have some nose down authority? And also, what happened regarding the THS? From the BEA report, the plane was trimmed up, but after the switch to alternate law. Who was trimming it? The pilots or the autopilot (which should have been disconnected...).

predictorM9; The elevators were max NU and held the attitude at +16°. I wouldn't get too excited about the stalled tailplane, as the drag caused by the vortex was assisting its lift, and airflow over the aerofoil, i.e. the THS and/or elevators, need only act as a lever to force the NU or ND as required.

Suggest you have a good look at some of the posts in the Tech Log to get an understanding of how the alpha protections and the associated autotrim work in Normal and Alternate Laws. Have a read of the Andy Tracy document at A340 - A330 Control: Flight & Laws, which is a summary of the basic flight laws for the A330 and A340.

Hi predictorM9,
Once stalled, pitch reached about 16 degrees, but AOA increased to above 60 degrees (about 61 at impact time). This change is due to the loss of horizontal speed, not to the THS or elevators position which only maintained the nose up attitude and killed most of the remaining airspeed due to added drag.
Who know exactly what kind of recovery was attempted? Nonetheless, it seems that some control (roll, pitch) could still be applied with some effects once stalled. We'll see that in detail when more DFDR data would be released.
Oscillations where registered before the stall and there isn't that much info about the later phase once she was effectively stalled, beside indications that the roll tendency was stabilized (likely meaning that ailerons imputs were still effective). The stall warning (SW) start well before entering a stall and it will even start earlier once an Alternate Law is triggered (due to the safe flight enveloppe reduction). Same about an overspeed warning (if airspeed was still computed) which would be reduced from Mach .86 to Mach .82. Despite the early SW (possibly due to pilot imputs or turbulences, hence g-induced in relation to AOA and Mach), she climbed 3,000 ft before effectively beginning to stall.
Alternate Law do not disable auto-trimming in manual pitch as it is only g-load related; consequently, it will follow any pilot pitch imput until an abnormal condition will be declared. On the other hand, FMGECs (flight computers used by autopilot) are not active anymore once ALT2 is triggered due to an unreliable airspeed situation: AP can not be re-engaged until this situation would be cleared (which imply that two ADRs must be valid following their logic).

Perhaps not really important in the overall context of your very informative post, but the airplane must have begun to stall shortly before reaching its apogee at FL380, and was well into a fully developed stall with AoA=16 degrees at less than one g when it got to FL380.

It seems most likely,finally, that the Pf was indeed the least experienced f/o ,in the rhs (frozen atpl)
Camel:
In my company the pilot in charge to replace the Captain's functions is the Senior F/O and his assigned position in the cockpit is CM2 (RHS). He is the Pilot Flying for the time the Captain is not on his seat. The "second" First/Officer will seat on the LHS to perform PNF actions, i.e., comms, papers and assistance to the PF. The "second" First/Officer is trained to execute several maneuvers (Emmergency Descent, Engine Failures, etc.) in order to be able to cope with any situation arising at a time when the First Officer may be absent for physiological reasons. He is otherwise never the pilot flying.

In the case of Air France, I am not sure of the practice or company policy, but I am sure they will have it written in their OM. (Remember that the word bureaucracy comes from a French word bureaucratie) :8

I tend to believe that the Pilot Flying would be the Senior First Officer and that he would be seated on his "natural", "most used to" RHS. Otherwise, I would say that, AF would have to change its policies.

Hi,

A doc ......

Interesting comment. When down elevator is applied what happens to the "vortex" drag that is "assisting it's lift"? Seems to me it would reduce.

Hi,

Nouvelles de Françe
News from France

L' interview du mois - Laengy Jo

I let for you the translation duty ......

Probably easier to to think of the "bigger barn-door" versus the "smaller barn-door". A bigger slab will be more effective than a smaller slab.

http://oi54.tinypic.com/pp2sz.jpg

A better understanding of what the above graphic represents may be obtained by viewing the initial graphic at post #1817 - AF447 Thread No.3. Also check out post #1786 in this thread.

EDIT :: The graphic has been changed to show a THS airfoil with bottom camber. Also, the local Angle of Incidence is shown.

Thanks for the diagrams and all of the other information.

However, I am still trying to figure out how any pilot would not realise that holding full backstick on anything other than an AN-2 (*) is a bad idea and is sure to get you into big trouble.

Is the full backstick recording actually true or is this an anomaly caused by some deeper issue ?. Did the PF *really* hold full backstick for almost 4 minutes, this seems hard to believe.

(*) The AN-2 is a special case, if the engine fails SOP is full backstick which pops the slats and causes the aircraft to mush slowly into the ground at parachute speed. - Only possible on a draggy old soviet STOL fossil.

nose goes up, wing stays down??
 

mm43;

i'm not a pilot - so maybe that's why your graphic confuses me: there's a 13 deg nu-angle depicted between "attitude" and "chord" - to me, this looks as if ' the nose would go up, while wing doesn't move with it' - should the chord not only differ from the attitude by about 3 deg?

Re aram
 

The THS (trimmable horizontal stabilator) was trimmed 13° to the aircraft nose up position. To get an aircraft nose up command, the THS is trimmed 13° down (leading edge of the THS down). The line "Attitude" basically represents the fuselage, and the zero position of the THS.

I think that should explain it.

thats exactly my understanding, RetiredF4 - now look at the graphic again: the 'fuselage' is in a 13 deg nose-up attitude, while the wing profile (chord line) is shown almost horizontally - imho, that can't go together - wouldn't the whole wing profile have to be shown pointing upward, in accordance with the 13 deg nose up attitude (apart from the small 3 deg difference between the two)?

EDIT: basic question: is the depicted profile representing the main wing, or just the horizontal stabilizer with trimmable section? - mayby that would explain my misunderstanding.

Quote from mm43:
The elevators were max NU and held the attitude at +16°.

Quote from The Ancient Geek
I am still trying to figure out how any pilot would not realise that holding full backstick on anything other than an AN-2 (*) is a bad idea and is sure to get you into big trouble.

As mm43 is now likely to be asleep, I'll offer my twopence-worth. The BEA does not say the PF had selected full back-stick, but you must remember that the FBW had only partially degraded to Pitch-Alternate Law – not Direct Law. (If it had degraded to Direct law, stick-to-elevator would have been in operation. But in Pitch-Alternate, that is not the way it works.)

Even a neutral (in pitch) stick would have resulted in full up-elevator, because the FBW would have still been trying to achieve (maintain) 1G, and the THS seems to have reached full nose-up trim (13NU) at around the apogee-point. Normally, Pitch-Alternate has a stall-protection function, based on AoA. But the AoA readings had been ruled invalid earlier because of the (false) low-airspeed data.

Although it was inevitably unsuccessful in maintaining 1G, and the aircraft's trajectory had become semi-ballistic, the system would have tried its best. In so doing, it would have selected full up-elevator. It no longer had recourse to the THS, which had reached full-travel.

Remember, the vertical-G that the system tries to achieve is governed by the position of the sidestick, unless protections come into operation.