A link from Evan...

[3WE]

http://www.bea.aero/docspa/2008/d-la081127.en/pdf/d-la081127.en.pdf

Outstanding link! Very interesting! The following comments are based on reading the report, starting at Page 18.

It would appear they pulled up relentlessly for several seconds (the report is not clear- 0 to 35 seconds of 'relentless' pull up and slowing)- but they did so with the intention of doing a deliberate stall (so no foul).

The warning and stall occur at about 99 knots.

The pilots immediately power up to take advantage of the big powerful engines which can generally power you out of an incipient stall with no altitude loss (although 99 knots sounds a little too late to power out of an incipient stall).

However, about 10 seconds pass before there's a nose down input.

During this time the plane displays traditionally ugly lateral stability and has big bank excursions with the pilot making corrective inputs with limited success (During this time the pilot even uses the rudder to make yaw corrections!!! How incompetent of him!).

The pilot finally goes full nose down and 5 seconds later, the plane appears to be 'flying' unstalled at 138 kts.

Then it's not clear what happens- perhaps they had already developed a significant sink rate and this resulted in a secondary stall- even though the airspeed and attitude "seemed" good (no huge surprise- that's what a secondary stall is- you think your attitude is healthy, but your AOA is not).

Anyway, 3 seconds later, the plane starts going "obscenely nose up" and this is in spite of the pilot giving further nose down input. From there, it seems that things go totally to hell.

I'm not sure what the failure was. As commented on another thread- airliners really aren't really designed to be stalled so who says they behave nicely when they do stall. I'd fault the guy for making this a 10 second stall at 3000 feet AGL- this thing isn't a Cessna 150. But it seems like he recovers ok...It's after that that bugs me some...the secondary stall...not unlike something I said to Gabriel in a PM...but how did the plane get itself to to those crazy nose-high attitudes with the pilot giving nose-down inputs?

[Evan]

The issue here, aside from deviating from the set test procedures due to ATC constraints, was frozen AoA vanes, apparently due to water ingress due to bad maintainence procedure. The AoA sensors froze at high altitude at a more-or-less cruise level AoA. WIth the AoA sensors frozen, the stall protection isn't going to work AND there will be no indication of this, no control law reversion... very dangerous situation.

They were testing the stall protections (demonstrating them actually) by slowing to alpha floor. They were not paying close attention to certain cues that would have told them something wasn't right (most noticeably the airspeed). They were completely confident and complacently relying on the systems. That's bad piloting.

When the a/c got into stall regime, roll excursions began at stall AoA and the pitots naturally began recording large assymetrical differences in airspeed, all three ADR's were subsequently rejected by the FCC and ELAC (although ADR 3 was not frozen and still accurate, hence the stall warnings), thus autoflight was lost and the control law degraded to Direct Law** (USE MAN PITCH TRIM message now displayed on the PFD).

** Direct Law will occur when the a/c degrades to Alternate Law and the gear are extended. The gear was extended in this case, so Direct Law.

As the a/c fell out at high AoA, the control law changed to Abnormal Attitude. The PFD appears now exactly the same as in Alternate Law, with no USE MAN PITCH TRIM message as in Direct Law. I see this as dangerous design oversight, but in extreme AoA and full stall recovery, I doubt any pilot would notice that message anyway)

In this condition. pitch trim MUST be done manually, The PF was apparently giving full nose down command but had also firewalled the thrust. This thrust, combined with the full ANU trim prevented the a/c from pitching down even with full forward stick.

Neither pilot recognized the problem or took steps to overcome it. That is the essential cause of the crash.

In any situation where the Airbus gets into prolonged stall warning (indicating that protections are lost), the pitch trim must be checked and adjusted manually. In any situation where the FCC protections are behaving erratically, two of the ADR's must be shut down to prevent erroneous AoA data and to defeat the protections and regain full authority. (Or, in the case of Quantas Flt 72, the entire ADIRU's, however that incident was a freak cosmic interference weakness that has since been designed out).

The point is, known your aircraft. Really know it.

Now here's what happens when you really know it: The reaction of this crew in identifying the issue and dealing with it is a model of professional Airbus piloting:

http://avherald.com/h?article=47d74074&opt=0

Thus far, two very dangerous, ice-related vunerabilities have been identified as compromising the redundancy of the Airbus systems. I hope these things can be overcome in design improvements but they must also be overcome in type certification training. In any case, the Airbus FBW is still fault passive if the crew is properly trained.

NOTE: I edited this to make a few corrections since posting it.

[Evan]

But it seems like he recovers ok...It's after that that bugs me some...the secondary stall...not unlike something I said to Gabriel in a PM...but how did the plane get itself to to those crazy nose-high attitudes with the pilot giving nose-down inputs?

Here's the rub...during the recovery attempt, the F/O called for gear up and the CPT confirmed gear up. Once the gear is retracted, control law reverts to Alternate again. Now pitch is no longer direct and is offset by the control law.

Position of the elevator in the last seconds From 15 h 46 min 00 s and until the end of the flight, the position of the Captain’s sidestick was at the pitch-up stop with a median transitional position of one second at 15 h 46 min 02 s. At the same time, the elevator position remained in a nose-down position.
Two factors can explain this phenomenon while the law in the longitudinal axis is a load factor law (the sidestick at the stop commands a normal load factor of 2 g):

• A rapid increase in the pitch is offset by the flight control law,

• A load factor higher than the value commanded leads to a nose-down movement of the elevator.

The CPT does not achieve nose down pitch until about 17 seconds before impact, at about 3000 ft ASL. Still, airspeed comes on fast and perhaps a recovery was still possible in Direct Law if the gear had stayed down.

Again, know your aircraft.

[Evan]

Here's where flying the plane comes in. Looking at the speed tapes below, it's plainly obvious that something isn't right. It shows alpha max at 70kts! Hello...

[Evan]

Please note that I edited my first post above to make some corrections.

[3WE]

Discussion appreciated.

...Gear...this mode...that Law...know your aircraft...go back and correct post for this nuance...

...I think if the plane had been maintained at proper airspeeds and attitudes using Mark IV eyeballs or even the much trained Flight attitude/status display it could have returned to healthy flight...

Works for: Boeing 747, Piper Cub, a whole #$itpot of airplanes in between. But with Airbus, there sure seems to be a lot of particular nit picky crap to keep straight.

Edit: WOW- so the great computer got tricked by full power and frozen AOA vanes and wouldn't let the pilot do basic airmanship things taught in the first two hours of flight instruction like lowerering the nose when in a stall situation.

...what a powerful example of why some folks object to the computer having too much "ability to think", over ride the pilots and fly the plane.

[Evan]

Edit: WOW- so the great computer got tricked by full power and frozen AOA vanes and wouldn't let the pilot do basic airmanship things taught in the first two hours of flight instruction like lowerering the nose when in a stall situation.

I don't see the part where it prevented the basic airmanship of watching the airspeed and not getting into a stall in the first place.

I admit that it is a bit more complex to recover from stall than a Boeing would be, although there's no reason to expect that a properly trained pilot, properly focused could not do it. The up side is that in non-upset situations it is a safer aircraft and that is 99.9999999999999999999999999% of the time.

Also, the XL scenario would never have happened if the paint crew had followed procedures, not unlike a certain 757 crash...

[3WE]

I don't see the part where it prevented the basic airmanship of watching the airspeed and not getting into a stall in the first place.

I admit that it is a bit more complex to recover.

Hey, we need to test the plane- let's relentlessly pull up with the deliberate objective of getting to a stall situation (the word "situation is important here) so we can test the stall protection system.

But the vanes are frozen and so we hold it a few seconds to long and actually stall the thing- Just like Boeing Bobby does to his 747...Just like Gabe does in his Tommahawk...no problem nose over...

Except, right out of 2001 Space Oddesy.

"What are you doing, Dave? You have selected full thrust and retracted gear Dave, that's consistent with a climb, Dave...your command for nose down is highly irregular, Dave. I'm sorry, Dave, nose down is not available, Dave...Close the pod bay doors, Dave..."

It's really sad as I imagine "Bicycle Built for Two" being sung while the damn thing wallows all over the sky in spite of a good early dose of nose-down input...

[Evan]

Hey, we need to test the plane- let's relentlessly pull up with the deliberate objective of getting to a stall situation (the word "situation is important here) so we can test the stall protection system.

But the vanes are frozen and so we hold it a few seconds to long and actually stall the thing- Just like Boeing Bobby does to his 747...Just like Gabe does in his Tommahawk...no problem nose over...

They were testing alpha floor, not 'getting into a stall situation'. If the apes who shot high pressure water into the AoA vanes hadn't done that, the alpha floor would have engaged well before a 'stall situation' with no need for recovery. They were not practicing stall recovery.

With those protections defeated by such idiocy, then it becomes more like the Boeing version or the Tomahawk version: watch your airspeed, reduce pitch on stall warning, DO NOT firewall the throttles (ok, maybe on the Tomahawk), except with the Airbus you must also tend to the pitch trim wheel because you are now in a 'situation' that simply doesn't exist unless you are a derelict pilot: specifically a gear-extended, Direct Law, fully developed stall that you deliberately flew yourself into, not by pulling up relentlessly, but by doing nothing at all. I seem to remember a certain Turkish 737 not faring well under this sort of piloting...

When the CPT entered full forward stick, he got the full downward elevator deflection he commanded. But the pitch trim was full nose up and, along with the underslung thrust, the elevators had little effect. Piloting 101, Boeing or otherwise.

They were around 118,000lbs, fully configured. On the A320 that equates to a stall speed of just under 100kts, therefore, alpha floor is going to be expected a good margin above that speed, around 107kts. Yet they flew it down to 99kts while (apparently) looking at a speed tape that put alpha floor at 70kts!!! Even a stripped-down A320 on vapors can't fly at 70kts! Any crew that can do this can also crash any Boeing or Tomahawk or Cessna ever made.

You constantly contradict yourself: you say it all comes down to piloting but then, when the problem is bad piloting, you say it's the airplane's fault. Does the reduced pilot workload on the A320 make it more conducive to pilot complacency? Yes, it probably does, but marginally because all airliners are now flown by automation most of the time. Can that be dealt with through training and screening. Absolutely. Should the Airbus architecture constantly evolve from these unforseen freak accidents. Absolutely. Or should we just give up and go backwards to the days of higher pilot workload and more fatal pilot error? Is that what you want?

I know by now that there's no point in arguing with you 3WE because you refuse to get your head out of the 1970's but this is the future, Airbus stall protections have saved untold numbers of lives and in addition to the basic airmanship required by all aircraft, pilots have to learn how to fly the planes of the age we live in. The second link I posted shows how that works in the real world.

[ATLcrew]

"It's a pity that everyone who knows how to run the government is already employed as a barber or a taxi driver"

-Francois Mitterand

[3WE]

"It's a pity that everyone who knows how to run the government is already employed as a barber or a taxi driver"

-Francois Mitterand

Valid.

I vow to not clutter PPRUNE and instead will stick to obscure photography fora for uninformed, unqualified, ass-hat parlour talk.

You're welcome.