Final Air France Report Out

[cncpc]

Nothing we didn't know, other than this...

...the wings gasped for air...

http://www.vancouversun.com/news/Final+ ... story.html

[pdw]

As if, with such rapidly developing ice (like what closed those pitots), there was a chance for recovery after stalling (plummetting).

Serious asthma attack.

[Colonel Sanders]

Not sure if people here understand or not, but it's
all about the money now. Many millions of dollars
in lawsuits is at stake here.

Air France desperately wants to deflect the blame
from it's pilots who didn't fly very well.

The company that made the faulty pitot tubes wants
to blame the Air France pilots.

Lots of lawyers are going to buy new BMW's on this.

[BEFAN5]

Maybe there was a English french barrier? The plane was screaming stall, stall but they thought they heard terrain, terrain, pull up.

[turbo-beaver]

And it urged an overhaul of the way France's aviation and airline industries are supervised.

Read more: http://www.vancouversun.com/news/Final+ ... z1znuIDTWh

And perhaps we should be asking ourselves.......is an overhaul of the way Canada's aviation and Airline industries are supervised also necessary?

I think, you bet. Air Canada is still playing silly buggar with their augment issues on the long haul flights and I hear Transport is on a hiring freeze and short of about a dozen inspectors.......so, who is watching the industry now?

[Siddley Hawker]

How can an aircraft be certified for flight in known ice with pitot tubes that are inadequately heated?
http://www.thedailybeast.com/articles/2 ... -down.html

[Eric Janson]

Not sure if people here understand or not, but it's all about the money now. Many millions of dollars in lawsuits is at stake here.

You are correct - it's a fight to determine who has to pay these millions.

I have seen a report of a similar event over the Pacific. The crew simply flew pitch and power and when the airspeed recovered they were within 5 knots of target speed.

I fly the same aircraft for a living - I still cannot understand why the crew did what they did.

All that was needed to prevent this was to maintain 2.5 degrees of pitch and to manually set a thrust of 80% N1. V/S and Altimeter indications would be normal. Flight Path Vector or "Bird" would also have been available. Put this on the horizon line and the aircraft flies level.

The aircraft remains very easy to fly even in alternate law.

[pdw]

The "Bird" is so useful for control ... and it's also shown that attitude was kept mainly straight/level in stall, right through the turbulence. The major nose down attitude for recovery is made late at 'realization' when the Captain comes back from his break ... from CVR/FR info. (I take it there was no other hindrance to stall recovery such as very serious icing ? Hard to imagine no airframe ice-catch with both pitots restricted by ice.)

Last night's news showed C.Sullenberger demonstrating similar controls, the 2 stick-yokes conflicting (CVR info). He makes a case for two control columns operating in unison (like the old way) in a crisis, and how those pilots were out on top of the storm at night (admitting it is a much different story ... from his own emergency in perfect WX).

[ogopogo]

So I'm just a part-time fair wx just-for-fun pilot, but here's my (stupid) question.... having understood the pitot icing, the stall, the apparent flight deck confusion, etc....why couldn't they recover from the stall? As one of the most practiced drills in private aviation, is this never practiced in the big leagues? Is stall recovery in the A330 not possible???

[YYZSaabGuy]

So I'm just a part-time fair wx just-for-fun pilot, but here's my (stupid) question.... having understood the pitot icing, the stall, the apparent flight deck confusion, etc....why couldn't they recover from the stall? As one of the most practiced drills in private aviation, is this never practiced in the big leagues? Is stall recovery in the A330 not possible???

Well...part-time fair wx just-for-fun pilot at this end as well, but I've done some reading on this point, so let me take a stab at answering your question. Hopefully one of the transport category pilots who visit this forum can straighten things out if I get too far off track.

Stall recovery (a subset of upset recovery/recovery from unusual attitudes) is included in sim training as a matter of course. It is absolutely possible to recover from a stall in an A330 - had Airbus not demonstrated the A330's ability to recover, the aircraft would never have been certified as airworthy. My understanding is that Airbus' test pilots (and Boeing's, for that matter) recover from a stall in the same way I would in my C172: lower the nose/release back pressure, accepting some height loss as inevitable, and power up.

As I understand it, though, line pilots are taught that altitude loss in controlled airspace is unacceptable (due to vertical separation requirements), and that stall recovery at altitude is achieved by powering up and maintaining a nose-high attitude to minimize/eliminate any altitude loss. Once the engines are spooled-up to 85%, back pressure is released to achieve a more appropriate attitude (5 degrees is the figure I've seen) and level flight achieved. If this "power first, attitude second" technique is used when the aircraft is thoroughly stalled, or is mishandled (by not reducing to a slight nose-up attitude once power-up is complete), stall recovery will be delayed. In fact, AF 447 was stalled almost continuously from FL380 to sea level, and was nose-high during almost the entirety of that descent at an AoA of up to 40 degrees (vs. a 15 degree pitch-up for a go-around).

The question is why the PF maintained the wrong attitude. The situation was complicated by the speed decay which resulted from the extreme angle of attack: as the stall progressed and airflow over the AoA and other sensors reduced, they could no longer generate reliable data and as a result discontinued their feed to the PFD, as they were programmed to do. Simultaneously, the stall warnings ceased. Once the AoA was reduced slightly and some speed regained, the AoA and other sensors resumed functioning and the aural stall warnings resumed, leading a by now thoroughly-disoriented PF to react by once again pulling back on the stick and increasing the AoA again.

So: in the face of everything that was happening on the AF 447 flight deck (finding themselves in the middle of a thunderstorm, unreliable speed indications due to frozen pitot tubes, numerous ECAM warnings, blanked instruments, repeated aural stall warnings, and a general loss of situational awareness all combining to heighten tension and make reasoning difficult), the two pilots flying did some of what they were trained to do to recover from a stall at altitude: they powered up and maintained a nose-high attitude. The problem was, they inexplicably remained very nose-high (evidently not believing they were in a stall) and as a result, lost control of the aircraft. What they failed to do, as I read the coverage, was trouble-shoot the larger problems (for example, diagnose why the stall warnings were being suspended) and figure out what their aircraft was actually doing. And not to reignite another Boeing/Airbus clash, but Airbus' sidestick design has also attracted comment as a possible contributing factor: they operate independently, which means the PNF wouldn't be made aware of control inputs by the PF because his own sidestick was designed to remain in neutral. Hence, the PNF wasn't aware until the very late stages of the descent that the PF had been maintaining a very nose-high attitude during most of the descent and that the aircraft as a result was stalled. By the time this became clear, they aircraft was at too low an altitude to recover.

Again, the above simply reflects my understanding of what happened based on my own reading, and I'd be very happy to have any Airbus pilots on this forum chime in to clean up any factual errors or misinterpretations I've made. But Ogopogo, to answer your original question: yes, it is absolutely possible to recover an A330 from a stall. Unfortunately, such a recovery wasn't achieved by the crew of AF 447 on that June night three years ago.

[Old fella]

"The aircraft remains very easy to fly even in alternate law"

I am not an airline pilot and certainly not qualified to comment on this particular incident as I have zilch experience or knowledge of this type of aircraft. However, I have heard the term alternate law and as a point of interest would like to know what it means. Is it a fancy term for what I understand in my old days as limited panel or is it much more complicated than that.

[Colonel Sanders]

http://www.airbusdriver.net/airbus_fltlaws.htm

[Flying Nutcracker]

It is hard to recover when you are not aware that the computer has switched modes, rendering YOU in direct control of the airplane without realizing the fact that the computer is not protecting the envelope. Having heard some detailed descriptions of the scenario at hand, explained by airline professionals, and not twisted in a media/press uneducated and sensationalistic manner, it isn't hard to see the complete confusion on the flight deck!

There was not a lot of communication going on between the 2 up front. When the captain finally showed up, it took a long time before the clue to the situation was discovered. By then, it was unfortunately too late.

The lawyers will have a good run with this one. For sure!!!

[Colonel Sanders]

It is hard to recover when you are not aware that the computer has switched modes

If Multiple Failures of Redundant Systems occur, the flight controls revert to Alternate Law. The ECAM displays the message: ALTN LAW: PROT LOST

Are Airbus pilots not allowed to look at the ECAM?

(for the humour-challenged - the above is a joke. It is intended for humourous purposes only. No other use of the above joke, including internal, is approved.)

[azimuthaviation]

Three hull loss incidents and a serious ground incursion in ten years for a European airline should not be something we're talking about in the 21st century, and besides air France, no one is. Hydroplaning on a wet runway, pitot tube failure, taxiing a large aircraft... These aren't new phenomenas, why ate they causing accidents? This is basic stuff isn't it?

[Beefitarian]

Maybe there was a English french barrier? The plane was screaming stall, stall but they thought they heard terrain, terrain, pull up.

Let's hope not. It was a French plane and airline!

[TheStig]

It is hard to recover when you are not aware that the computer has switched modes

For all the curious non-airbus types: Anytime the aural warning 'STALL' is heard, the A330 is not in 'normal mode'. It will not stall in 'normal mode'.

[2R]

Was the RADAR working ?
I would guess it was not working at 100 percent,why else would any pilot fly into a large nasty CB.
Maybe they should take a look at the RADARs on their planes as this was the second fight they lost with a CB.

[TheStig]

Was the RADAR working ?
I would guess it was not working at 100 percent,why else would any pilot fly into a large nasty CB.
Maybe they should take a look at the RADARs on their planes as this was the second fight they lost with a CB.

From what I've read this flight had been filed to Madrid with CDG as an alternate. This procedure is known as In-flight re-fileing or re-dispatch. It is becoming increasingly common as commercial pressures to minimize costs mount. Unfortunately, carrying extra gas does not generate extra revenue.

Typically, as a re-dispatch flight approaches Europe (or North America flying westbound) the pilots and dispatch re-check the destination weather and the remaining fuel to determine whether it is safe to continue too the planned destination or divert to the flight planned filed destination.

It is a safe system, however, it places extra pressure on the pilots to cross the ocean without burning into to the fuel require to make it to their planned destination. We all know how much pressure 'get-home-itis' can put on a flight crew. I'm not trying to speculate on the cause of this crash, but don't you think the crew of AF447 would have diverted around the thunderstorms if they knew they had the fuel to do so and continue to CDG?

This is an excellent youtube video;
http://www.youtube.com/watch?v=ARybu2kH ... ata_player

[sidestick stirrer]

I recall reading that the pilot occupying the left seat, just before the incident began, had only recently relieved the Captain and, in conducting his " coming on duty" panel sweep, noticed that the weather radar settings were not correct.
Once the gain and tilt were returned to an optimum setting, the display showed a large return right in front of them requiring an immediate turn away. With the speed of the aircraft, they were unable to avoid penetrating the cell(s).
I think that was the first hole in the first slice of Swiss cheese (Reasoner) to begin the error chain. The next was the experience level of the PF( six years, IIRC) who obviously remembered that just holding full-aft sidestick would allow the magic to fly the aircraft out of any nasty situation.
In normal law.
With envelope protection.
Neither of which the airplane was enjoying after the magic said," not enough information. I give up, it's all yours".
The third hole to line up was the overwhelming amount of information the pilots had to decipher. Hidden behind all the tombstones and birds and mustaches and cues was the attitude sphere, and to add to the din the supercooled water droplets must have been making as they hammered the nose was multiple aural warnings, cautions, the altitude C-chord, autopilot disconnect warning, the lot. The warnings and cautions were popping up so fast on the upper ECAM, each new one either driving all the others down the list or rearranging the list as the magic tried to prioritize each new one relative to everything already listed that it is impossible to even read a two-to-three word item before it is snatched away, to reappear somewhere else.
The PFD alone displays 157 different bits of information, all at the same time.
Think about that number for a moment.
How to sort the wheat from the chaff when everything is shouting and flashing at you for your attention and none of it makes any sense?
Why, revert to basics and what you know to be true about your plane. Hold the stick full back and the magic will take care of everything. It will pitch up to the optimum AOA, reducing the sink and be at the best attitude to fly out of anything.
Now, it is true that this also happened to 340 crew shortly thereafter only they turned all the magic off, erased all the distractions from their PFDs and flew attitude and power until getting sorted out. They may have been more experienced, may have been flying in daytime and may not have blundered into a CB.
The last slice of cheese with a hole that led straight to the crash was the sidestick controllers. As my forum name implies, I have years of experience with them ,stirring the pot, as it were.
They work just fine. The only real difference is that-in the air-there is no visual indication of what the PF is doing with his sidestick( on the ground, there is the iron cross that displays the summed input of both sidesticks). The more-experienced pilot in the left seat did not realize that the nose-high attitudes were being caused by the PF holding his stick full aft.
It was not until the Captain returned and, standing behind the two, was able to see, in the dark, that the PF had full back stick. He then ordered him to pitch down, which he did.
The aircraft began flying again almost immediately but struck the ocean while still recovering from the ensuing dive.
So, there is lots to glean from this and- if we don't-then it will most-assuredly happen again.
When we are overwhelmed with too much information or distractions to be able to sort it all out, your house of cards has fallen down and needs rebuilding. The foundations of this house are what your instructor tried so hard to teach you: attitude plus power must equal performance.
Just as there are laws of physics that the airplane can't overcome, neither can it ignore certain combinations of attitude and power; it MUST perform in response, it has no choice, windshear encounters excepted.
Flight controls that move on both sides of the cockpit are worth the weight penalty and complexity.
The magic is nice but can be a two-edged sword. Routes and approaches are now built without a single waypoint related to a terrestrial navaid. It also lulls us into complacency: who bothers to figure their own top of descent points on an FMS/FMC-equipped airplane anymore?
This entire event could have been broken anywhere along the chain.
A simple change in the display logic would allow the iron cross to appear in extreme situations so that the PM can see what the PF is doing with the controls despite the obvious issue that it would be yet another thing cluttering up the attitude sphere.
Finally, although there is no going back at this stage, I rue the day they began building these big things without a third guy in the cockpit. Who has ever sat in the "Captain Kirk" seat in a simulator or ridden the jumpseat and not had both a grand view of the overall picture but was removed just enough from the needle-chasing and stirring to have the best grasp of the situation? Standing just far enough from the trees to be able to see the forest? FE's and S/O's saved untold incidents and accidents from happening but unfortunately they ate food and cost jet fuel to carry around and needed hotel rooms and pay cheques and a pension and a...where was I?

[pdw]

With any ice on the wings it is also very possible there wasn't the chance to recover once the point of stall was reached near full gross weight. It says minus 40C, however before the ice shut down the pitots the copilot explains "ozone smell" and comments on the sudden heat in the cabin (CVR info).

The pitots are heated and operational very soon after but may only have been possible for attitude/speed recovery right after sudden climb from 35,000 to 38,000. Once in stall with ice it is not the same. Even clean, how many such stall recoveries are on record ?

[flyinthebug]

Sidestick...great rant!

I agree with almost every word you typed. I strongly agree with your nod to the S/O`s that should exist in every commercial airliner. If AF 447 had an S/O onboard, he/she would have detected the incorrect inputs to the controls and advised long before the Captain came back and noted the problem (a bit too late). What if an S/O was onboard and noted this problem 30-45 secs sooner?

I also agree with your assessment that they should have simply went back to basics(attitude+power)...shut down the ECAM and just flew the plane. I know how easy it is to armchair this one, but some simple basic flying would have avoided this entire tragic accident, despite the pitot failures.

Once again, great rant. I sincerely think you nailed it.
Fly safe all.

[sidestick stirrer]

Thank you for your kind comments.
When I started my career, as a S/O on DC-8's, in the spring of 1973, I had less-than 275 hours TT, so was fully prepared to just sit quietly on my hands and learn from the Sky Gods I was about to fly with.
No question, they were real professionals , extremely talented and also very kind to The 22-year old kid sitting behind them.
But they were also human.
In the eight months I crewed The Mighty Eight, before moving on to the Classic, just off the top of my head, I can recall them:
Forgetting to put the gear down( more than once)
Forgetting to extend the flaps from approach to landing
Forgetting to arm the spoilers
Forgetting to set the trim before takeoff
Setting the wrong flap for takeoff
Forgetting to tune a nav radio for a SID or approach
Not detecting a thirty-degree error in heading from a failing flux valve
Many missed items in procedural flows
Wrong frequencies dialled up
Reminder bugs misset
Many misunderstood or not overheard ATC calls

While this list may make one shake their head at what seems like gross incompetence, few realize what level of technology these first-generation jet transports came with, like the first flight directors ( a single, slow-moving, vertical steering cue that waved from side to side) and AM-only HF radios, single DME, no ground speed readouts, very-few warning lights or aural warnings, etc. Coupled with many of the destinations being ESL and arriving there in the wee hours of the morning, body time, after having spent at least seven hours sitting in darkness, it was amazing that it worked as well as it did. There was no such thing as a relief pilot or controlled rest back then.

As I said, these pilots were the best that I had ever or would ever have the pleasure to work with but, once I adjusted to the fact that this was just basic human fallibility, then I noticed that none of them made light of, downplayed or criticized the other pilot. It was "Oops" or "Sorry" must most-often "Thank you" when I spoke up upon noticing these things. That made them look even better in my mind, as they obviously appreciated my help and welcomed this wet-behind-the-ears kid who couldn't spell Mach tuck or tropopause let alone explain them.
True professionals, most of whom had flown Spitfires or Mosquitoes or Lancasters during WW2 and were now in the last years of their careers. That was the Captains, nearly all the First Officers had flown Sabres or CF-100's in Europe.
I emulated what I saw and heard right through my fourty-one year career, and I could visibly see my new-hire Cruise Relief Pilots relax when, just before pushback when we were about to review our emergency drills, I would turn around to them and say that I was glad there were here, that I was human, was getting dumber with each passing day, that I stick as close to SOP's as the situation allows, and please don't ever hesitate to speak up if I miss something, if something doesn't make sense or if they see something developing that I don't. Many of the F/O's would also add that went for them, too.
Much like a canned PA to the customers, I always said the same thing but it never once failed to create the working environment in our flight deck that I felt was optimum.
I had sat back there where they were, long enough to realize the valuable overview and insight that unique location allowed and I wanted and needed that third set of eyes and ears.

Warmest Regards,
Scott

[Eric Janson]

So I'm just a part-time fair wx just-for-fun pilot, but here's my (stupid) question.... having understood the pitot icing, the stall, the apparent flight deck confusion, etc....why couldn't they recover from the stall? As one of the most practiced drills in private aviation, is this never practiced in the big leagues? Is stall recovery in the A330 not possible???

It is normally trained during initial Type Training and then never done again. We had it on the last recurrent because the procedure has been changed.

If you've stalled a transport category jet then you've done something very wrong imho.

Simulators are far from perfect - there are issues with the accuracy and realism of the modelling used.

I am not an airline pilot and certainly not qualified to comment on this particular incident as I have zilch experience or knowledge of this type of aircraft. However, I have heard the term alternate law and as a point of interest would like to know what it means. Is it a fancy term for what I understand in my old days as limited panel or is it much more complicated than that.

In very simple terms

Normal Law - The aircraft has limits on speed/pitch/roll and load factor which cannot be exceeded.
Alternate Law - Most protections are removed. The aircraft can stall. Activation of Alternate Law requires multiple failures.
Direct Law - Aircraft flies like a C-172 and needs to be manually trimmed.

The important thing to note is that in all cases the aircraft remains perfectly flyable.

I think some of you need to spend some time reading the report.

http://www.bea.aero/docspa/2009/f-cp090 ... 601.en.pdf

From page 17.

The accident resulted from the following succession of events:

-Temporary inconsistency between the measured airspeeds, likely following the obstruction of the Pitot probes by ice crystals that led in particular to autopilot disconnection and a reconfiguration to alternate law,
-ˆˆInappropriate control inputs that destabilized the flight path,
ˆˆ-The crew not making the connection between the loss of indicated airspeeds and the appropriate procedure,
-ˆˆThe PNF’s late identification of the deviation in the flight path and insufficient correction by the PF,
ˆˆ-The crew not identifying the approach to stall, the lack of an immediate reaction on its part and exit from the flight envelope,
ˆˆ-The crew’s failure to diagnose of the stall situation and, consequently, the lack ofany actions that would have made recovery possible.

[xysn]

Why, revert to basics and what you know to be true about your plane. Hold the stick full back and the magic will take care of everything. It will pitch up to the optimum AOA, reducing the sink and be at the best attitude to fly out of anything.

That's a very interesting observation. We all have fall-back thinking, but to have THAT as fall-back highlights an issue with training, I think.

My own questions - answers to which may never be known:

Why does Bonin apply an initial nose up input? Was he trying to fly over the storm?

Why does the Airbus "dual input" system "average out" the inputs? Why does it accept two simultaneous inputs at all? It's not like with FBW that you need additional muscle power to overcome pressure on the control surfaces.

Although Robert asserted control twice, Bonin retook the controls on both occasions to apply nose-up. Why? It seems terribly basic to (I learned this right after learning how to fasten the seatbelt ... in a simulator) be clear on who is flying the plane.