Would you set TOGA if you believed to be in an overspeed scenario?

Assuming they still had correct pitch information (and there's no indication whatsoever that they hadn't) it's hard to understand how they might have come to that conclusion.

This leaves
which doesn't sound that good either :uhoh:

AF 447
 

As one whose very limited pilot hours were a long time ago, might I ask if any simulator time is spent on flying difficult manoeuvres on basic flying instruments, Gyro driven horizon, attitude indicator etc (presumably basic turn and bank is no longer there.
The ability to realise what was happening seems to have been a factor.

Yes, this is not a new occurrence for Airbus or Boeing.

I've done the unreliable airspeed training in the sim from 10,000' down to landing. There are a few things you need to think about ( obviously !! ) but it's manageable without any Airspeed indicators. Granted it was in a Sim, reasonably good wx and with no risk of dying ( my heart rate was relatively normal in the sim !!) but we did have a lot of bells and whistles going off for a while which proved a headache for the PM. My job wasn't overly difficult, fly the power and attitudes for the config etc.....

Every Pilot in my Airline did this as part of their 6 monthly recurrent training package.

Stu B, The 747 uses the nose wheel squat switch, the stick shaker goes off when the A/C is jacked if the CBs aren't correctly pulled prior to jacking.

At 2 h 10 min 51 the stall warning was triggered again. The thrust levers were positioned in the TO/GA detent and the PF maintained nose-up inputs. The recorded angle of attack, of around 6 degrees at the triggering of the stall warning, continued to increase. The trimmable horizontal stabilizer (THS) passed from 3 to 13 degrees nose-up in about 1 minute and remained in the latter position until the end of the flight."


Is it normal for the stall warning to be triggered at 6 degrees ? To the uneducated this appears quite low. Wouldn't the stall AOA in clean config be around 13-15 degrees?

If the aircraft was still in Alternate Law as indicated by PNF is there anyway the automatics could trigger an uncommanded (by the pilots) THS nose-up change of 10 degrees?

Can anything be read into the rate of increase in nose up ? To the uneducated it appears relatively slow / deliberate ?

Agreed 100%.

I suspect the scenario will become clearer once we know what pitch information the pilots were seeing.

Also, note that the speed information displayed to RH pilot is not recorded. The PF was in the RHS. The PF may therefore have been reacting to speed information different to that recorded.

I think it is interesting to revisit the situation the Captain William Hagan on the BA 744 (LGW-NBO) in December 2000, who came onto the flight deck to find an intruder attacking his F/O and the very dangerous stall that the aircraft was in. Another four or five seconds he said, and the aircraft would have been on its back! Again it was at night with no horizon and credit must be given again to that F/O who saved the aircraft without overstressing it. I would say that was a worse predicament for the BA 747-436 than the AF A-330.

Incidentally I flew that 744 to PHX soon after the event with our new CEO, Rod Eddington, on board. My F/O was shocked to find there was still blood spattered on some of the instruments! Eddington told me that he had 'phoned the President of Boeing to congratulate them on building such a strong airplane.

Gentlemen, it is all down to airmanship. Reading the reports in the Press of Air France congratulating the heroic efforts and professionalism of the flight crew on AF 447, makes me think if you are going to tell a lie, you might as well tell a whopper.

In a similar situation on a 747-236, my instinct was not to raise the nose and put the aircraft even closer to coffin corner. The AF pilot(s) allowed the situation to deteriorate where the aircraft had almost no forward speed or momentum. How do you recover from that?

The fact that Air Caraibe pilots demonstrated the appropriate airmanship in two similar scenarios, shows that basic flying skills are lacking at AF. How can you otherwise explain landing an A-340 half way down the runway in a thunderstorm at YYZ and expecting to stop, the heavy landing at CCS, the taxying incident at JFK and even the loss of the Concorde may have been a different result, if BA procedures had been followed as well as maintenance. Maintenance of course, regarding the pitot tubes (changed by Air Caraibe following their upsets) is again a factor in an AF crash.

Air France needs to go through a flight safety audit similar to Korean Airlines, who since that have not suffered a hull loss.

C.N. you sure got that right!

That sounds like the missing procedure, and hopefully one that will be thought hereafter!

As I said above this scenario is not new unfortunately. There are now quite a few documented cases around the world with Boeing and Airbus operators. Enough incidents that Boeing and Airbus have for some time now included checklists in their QRH's and also recommended all crew be exposed to it in the Sim during their recurrent training cycles.

Thread drift.........This maybe more proof that employing low time inexperienced crews is just plain dumb, Airline managers trying to save money please take note.:=

There is no substitute for experience, full stop end of story:ok:

As expected reading the last few pages, the usual crowd of perfect pilots comes on to disparage those pilots who were imperfect, the only difference between the perfect pilots and the rest are that the perfect pilots strut around here like a bunch of cockerels, the imperfect pilots are out there facing the real world scenarios...

Anyway that aside, to me what was said and the actions of the PF, with no real dissent from either of the other two imperfect pilots who were there, indicate that something either in the instrumentation or the plane's response to inputs was diametrically opposed to what the crew expected to see or experience. Unless you perfect pilots believe that all three highly trained pilots on the flight deck on that dark night were so grossly incompetent that they could not follow basic airmanship as a matter of course, then there must be some factor(s) that the data recorders have not been able to provide and which may never be known.

As for the dogmatic statements by some perfect pilots above that the crew shouldn't have flown into the storm or flew a perfectly serviceable aircraft into the sea, your comments are beneath contempt, not only for their insensitivity but also for their gross simplification of a situation that you in reality know absolutely nothing about.

Could someone explain to me why there was no distress call in 3 minutes ?

An interesting read, forget technology and stall recovery techniques, I believe beyond the system failure the cause to this accident runs deeper. Whenever we go into a degraded mode the assumption made is that the pilots have the skill recency to deal with the problem. My view is that modern EFIS systems have led to pilots forgeting to scan their instruments. Their focus is on the MCP/FCU and the FMAs not the attitude or performance instruments. Even on the approach rather than looking at the speed many pilots just trust the autothrust to maintain it.

Retaining basic instrument skills is easy even with a low tech PC flight simulator but few people do it unless they are going for an interview. Until this issue is addressed, events such as this will continue albeit infrequently.

Power + Attitude does equal performance but first you have to look at the instruments.:(

No AoA indication?
 

The key question here is - 'did the pilots have enough information to realise they were stalled, pitched up, throttles closed, and recover'?

I am surprised that in all this discussion (including that of the thoughtful Shadow) putting forward idea that the pilots didn't know what the AoA is, nobody has mentioned the backup AI (about 6 inches to the right of the captains MFD)? Was it really not noticed by 3 experienced pilots that the there was a huge difference between PFD and backup AI, which would have indicated a big pitch up, and yes (as mentioned by another poster) - no noise in the flight deck? Isn't this why the backup instruments are there? Or am I missing something (only a humble PPL/IR)?

Obviously there's a fundamental difference between AoA and attitude (ie the vertical speed of the air) but I am trying to think of a conceivable Wx scenario that would deliver up to 40 degrees pitch up on the AI with closed throttles for an extended period, and not indicate a stall.

nitpicker330 - yes! Power+Attitude=Performance. They had power - they had a backup AI. They also had, I presume, an 'alternate air' feed (like my Mooney!) which would have got the altimeters working again, at least the backup ... There was a lack of training and/or airmanship here, probably a huge dollop of denial (surely EVERYTHING can't have gone wrong!), though it looks as if Airbus need to seriously look at some kind of mechanism to warn the pilots if the computer lacks confidence in it's view of the world (due to inconsistent inputs) and advise the pilots to check the backup instruments.

We have all been trained and examined in the effects of icing in the instrumentation. It seems amazing that 3 experienced pilots wouldn't have guessed something was up, especially IR onwards.

All in all, I am concerned that it can't be as portayed as those 3 guys should have figured out what was happening.

Never a truer word...

Capt was away from the cockpit resting!
 

Whenever flying as multiple/augmented crew, my company schedules 2 captains & 1 F/O (this was gotten forcefully accepted by our pilots' association and the captains). So at all times you have 1 capt on the flight deck and somestimes 2 capts whilst the F/O is on a break. In the relaxed cockpit crew scheduling regime at AF and possibly other reputable carriers as well they put 2 F/Os in a 3 member set. Considering enroute WX forecasts and/or actual WX returns the Capt on AF 447 should have retained his seat and reclined it for rest. I can't understand how any responsible capt would prefer his allotted rest over safety of operation especially if that rest meant being away from the flight deck. Sheer negligence I'd say. AF and other companies that allow crew rest periods need to urgently review their scheduling policy and advise their crew to exercise sound judgement and discretion for rest periods considering segments of bad enroute WX. I always thought that one takes adequate rest prior to assuming duty and layover time is best spent and meant to arrive fresh for duty. I hope and am sure that the final report will adequately address the question of this capt, his rest break, and absence from the cockpit.

Yes but it's not only the Pilots that need to do this.
The onus should also be on the operators to 1/ recruit the appropriately experienced crews and 2/ ensure they retain their Piloting skills through targeted appropriate training packages throughout their career insisting they maintain basic instrument skills and know how to fly raw data ( this is not encouraged at all, indeed it is frowned apron now days )

I have a question:

the news media has described the 3 minute drop into the ocean as having been 'horrifying' for the passengers. However, if indeed the pilots didn't even realize that the plane was stalled and was descending so quickly, could it be that the passengers were completely unaware of anything being wrong during their final minutes?

You can only feel a fall during the initial acceleration. And surely if the pilots could feel that 10 000fpm decent then they would know they must be stalling and stop the nose up attitude inputs.

Hi,

I think they were so busy that if you would ask them their names they would not have been able to respond

Fatal flaw...... It's my understanding the ALL the Attitude indicators where fully functioning all of the time throughout. It was only Airspeed that gave incorrect indications via iced up Pitot tubes screwing the 3 ADM's.

nitpicker330 - I guess you must be right otherwise the flight data recorder wouldn't have recorded it ... so they spent 3 minutes pitched up with the throttles closed and the altimeter plunging waiting for the speed to come back down ... what did they think was happening? :uhoh:

In all the documented incidents so far there has never been an issue with the Inertial Attitude displayed on the primary flight displays or on the separate standby. A simple cross check of all 3 would confirm the same.
Airspeed issues and associated overspeed or underspeed ( or indeed both ) warnings were what happened requiring the crew to look "through" their displays and revert to "basic flying skills"

Oh, and no modern western built transport has an AOA indicator fitted.

Who knows what this crew were thinking, who knows how fatigued they were on the night? Who knows how much Turbulence they were dealing with?

All things being equal they "should" have been able to deal with the situation, obviously all things weren't equal.:(

nojwod Is there any explanation/reason for the two momentary roll inputs by the PF?
Could these indicate a deliberate check by the PF as in, 'Am I really connected to the control surfaces?'

my comments as a PPL/IMC holder:

If I know that Power plus attitude=performance, so did all 3 pilots there that night. They had power (the engines were working normally) they had attitude info (attitude indicator, not derived from pitots) so obtaining a performance seems the logical way to go. The time the descent took rules out the startle factor, and as is well known, once the descent stabilises at 1g, the seat of the pants feel is useless at best and misleading at worst, yet they still should have had the AI. Again, even as a PPL I know to trust my instruments and ignore the seat of pants sensations. So, why didn't they??
I can see a couple of possibilities- 1. they didn't trust the AI as they were aware of the pitot problems and decided to go with the seat of pants sensations or 2. the stall they found themselves in was irrecoverable due to some design problem or issue inherent in the jets design. No comment whatsoever on AB vs boeing, but don't swept wing jets has a nasty characteristic of entering a flat stall which is hard/very hard to break?
Given the possibility of option 2, and if I were AF or AB I would be looking to promote option (1) since option (2) has the unpalatable effect of admitting that some stalls are invariably fatal, not something which passengers want to hear. So much cleaner to blame it on human error, psychologically, its easier on passengers to believe that the pilots made a mistake, rather than face the possibility that a particular set of circumstances will result in an invariably fatal outcome. What do the stable extroverts amongst us think?

Nitpicker330, (post 700) sorry but I have to call you on that statement. Many Boeing 777's and 737NG's have AOA indicators displayed in the upper right corner of the PFD. Delta Air Lines and American Airlines started this trend back around 2001. It is a Boeing option, you just have to tick the box when you order the airplane and they will deliver the AOA. I believe it's standard on the 787.

Fair enough I stand corrected :ok:

The info is in the system so it obviously isn't too hard to program onto the display. Maybe it will become a mandatory update for all soon?

Why did the autopilot then auto-thrust disengaged ?
 

"From 2 h 10 min 05, the autopilot then auto-thrust disengaged and the PF said "I have the controls". The airplane began to roll to the right and the PF made a left nose-up input. The stall "

At this time, it seems that there were no Pitot problem ?

Why did the autopilot then auto-thrust disengaged ?

Yes, there was. That's exactly why it disengaged.

Similar erroneous AOA indication experience on B767300
 

A couple years ago, I had to fly between two mounstrous squall lines (30nm appart) on a B767300ER. I was descending thru FL250 in dusty air (no water content) for an approach into Lahore, Pakistan, when the stick shaker was triggered continuously for about 1 minute with severe turbulence. I was on FLCH, thrust idle and the airplane was climbing at 500fpm for the whole minute. The AI was showing about -5 to 0 degree. Until today, I assume I encountered the mother of all updraft and for that reason, a high angle of attack information was sensed and triggered the stall warning system.

I was then transferred to the Airbus fleet and prayed everyday I wouldn't encountered the same scenario aboard the A320 I was flying because I had no clue how it would behave in such a situation. I am sure it would be way more confusing than how it happened in the 76 due to the relative complexity of the flight envelope protection system.

This is not an Airbus bashing post but a personal experience that may help understand what happened in AF447.

If it ain't Boeing I'll be going but not by choice...

OAT
 

Is any OAT information available? I am recalling the Pulkovo 612 accident, that beside a daredevil decision by the Captain to overclimb CB also was hit by a monstrous updraft sending the plane into a 16,000 fpm zoom to FL420, where it stalled.

"Pulkovo 612 departed Anapa (AAQ) for St. Petersburg (LED). The Tupolev climbed to the cruise altitude of 35,100 feet (10.700 m). Because of storm cells ahead, the pilot decided to change course laterally by 20 km and attempted to climb over the storm cells. However, the thunderstorm front was unusually high, extending up to 15 km (49,000 feet). The Tu-154 entered an area of severe turbulence, pushing up the airplane from 11.961 m to 12.794 m within just 10 seconds. The angle of attack increased to 46 degrees and the airspeed dropped to zero."

Very little is being said in this thread about highly unusual and unexpected WX phenomena that may have caused the rapid freezing of the pitots and the initial upset. Could sudden changes in OAT those minutes into the flight give a hint?

Pardon me for entering ID badge area with my RTF and ICAO 5 English rating.

Auto trim is disabled in Alternate law
 

Quote:

---

I too am gobsmacked, that in this day and age Airbus have managed to design another generation of machine that doesn't do what you could reasonably& logically expect.

---

This is not correct. When the system is in alternate law the auto trim function is disabled. Pushing the stick forward only results in limited elevator motion. Trim must be addressed by the pilot in this mode, so simply pushing the nose down isn’t going to retrim, the pilot has to do it manually.

This still doesn’t address why the trim was moved to the 13 degree nose up position.

@Zorin_75
 

I have not read in the BEA report that the Pitot had trouble before the first stall alarm ?

The QRH spells the situation quite clearly:---

Unreliable speed indic/ADR check proc:-

Maybe due to Radome damage, air probe failure or obstruction
Indicated Alt may be effected if static probes effected
Unreliable airspeed cannot be detected by the ADIRU

Since Flight control laws maybe effected maneuver with care
Unreliable speed may be suspected by-
--- Speed discrepancies between ADR 1, 2, 3 and standby
---Fluctuating or unexpected increase/decrease/steady indicated speed or pressure altitude
---ABNORMAL CORRELATION OF THE BASIC FLIGHT PARAMETERS
---Abnormal AP/FD/ATHR behavior
---Stall warnings, or overspeed warning or flap relief warnings that contradicts with at least one of the indicated speeds
-RELY ON THE STALL WARNING THAT COULD BE TRIGGERED IN ALTERNATE OR DIRECT LAW. IT IS NOT EFFECTED BY UNRELIABLE AIRSPEEDS BECAUSE IT IS BASED ON AOA
-DEPENDING ON THE FAILURE, THE OVERSPEED WARNING MAY BE FALSE OR JUSTIFIED. BUFFET ASSOCIATED WITH THE OVERSPEED VFE WARNING IS A SYMPTOM OF A REAL OVERSPEED CONDITION.
---Inconsistencies between radio altitude and pressure altitude
---Reduction in aerodynamic noise with increasing airspeed or increase in aerodynamic noise with decreasing speed

*my capitals to emphasize some sections*

From the Grauniad. Surely there would be multiple pitots. Could there be a software problem in "voting" in event of discrepancies??

A pilot's analysis

You look at this as an A330 pilot and it's a case of there but for the grace of God. The first sign of trouble was the expectation of turbulence voiced by the first officer. He reduced the speed in anticipation to Mach 0.82, which is normal, and warned the crew. Shortly after entering the turbulence and associated ice, the autopilot disengaged and the first officer commented that the flight control computers had disconnected. This happened because the computers were no longer receiving speed information and neither were the pilots.

I cannot speculate why the pilot climbed the aircraft but it could have been due to confusion caused by the turbulence and sudden instrument failure. The controls at very high altitude are very sensitive and there is only a very small speed envelope, often called "coffin corner", where only a few knots of speed up or down can cause a stall.

So you are flying in this little window of a flight envelope which is perfectly normal and, indeed, you do it every day. But if you are suddenly handed control of an aircraft you can breathe on that little stick and the aircraft can go down one degree. In a very short space of time lost speed and stalled, causing it to fall from the sky. Despite what the pilots must have been going through in the cockpit, the pitch attitude remained fairly constant throughout so the passengers would not have sensed something was seriously wrong. They would have felt mild buffeting of the stalled airflow over the wings and the initial sinking feeling. That's all.

I feel a lot of empathy for the pilots. One could imagine suddenly suffering a total loss of airspeed indication in turbulence in the middle of the night and with all the autopilot systems suddenly failing. The indications, both visual and aural, would have been very confusing. The question for me is why did the instruments that are triple redundant, which means that there are three back-ups, all fail together?

Anonymous A330 pilot

Repeat after me
 

Every mother or wife of a pilot should say every day to her son/husband :

"repeat after me one hundred times : when I hear a stall warning I nose down..."

There is absolutely no reason why the PF :

"At 2 h 10 min 51, the stall warning was triggered again. The thrust levers were positioned in the TO/GA detent and the PF maintained nose-up inputs. The recorded angle of attack, of around 6 degrees at the triggering of the stall warning, continued to increase. "

This is unbelevable exept if you want to suicide.

Hi,

The dinosaurs must think I'm sure the good old days when you had to put the feet on the dashboard for manoeuver elevators to get out of a dive
In the case of FBW aircraft (Airbus particulary) .. may have the same maneuver with two fingers without any effort.
It certainly does not help to understand what is the force exerted on the moving parts

HKPAX:-- Please read my post above. Or better still read the actual QRH you have.
All or 2 of the Pitot probes iced up ( basically failed ) and "confused the system" Overspeed/Under speed warnings resulted, Alt Law, AP A/THR dropped out ......etc etc....

It would have all cleared up after they exited the cloud........

Airbus and Boeing have addressed this issue of super cooled water drops and directed that improved pitot probes be fitted.

Altitude and lateral movement
 

Would anyone like to comment on this:-

The aircraft is telling the pilots (probably correctly) that they are at 10,000ft. The quoted text implies that the pilots believe this. What mechanism (stall/dive/something else) do they believe has got them from FL350 before any problem arose to FD100 (I know they ended up at one point at FL380 but we don't know that they knew that)?

What manoeuvre are the pilots trying to make on the sidesticks and is there any reason why both of them would be making simultaneous inputs other than that they do not trust one or both of the controls to react to the input?

AF Atlantic
 

Gentlemen & Ladies of course,
Having watched the story since the beginning, I think it is time a few salient points should be pointed out. I think I am qualified to comment having spend 5000 hours as P1 on 330's although I am now retired. I would therefore be happy to accept any corrections.
As I recall from the original ACARS transcripts, all pitot statics were iced up.
The effect of this disconnects the A/P and A/T without pilot input. The airspeed on both PFD's and STBY ASI would be useless and in total conflict to each other. 2 warnings follow, STALL STALL STALL(verbal) followed by immediately by the OVERSPEED warning (I think it is the "Cavalry Charge" bell).
Any sidestick input at this point would be extremely fraught (remember the autopilot is now off).
So, which one do you believe. It seems like the crew were trying to recover from an overspeed event and ended up stalling (throttles closed, pitched up).

The early analysis of the wx showed the flt to be transiting the ITCZ. I spent more hours transiting the Asian one to know that it is not to be taken lightly.
The question then begs-
was the crew aware of the weather approaching;
was the weather radar operating; was it being operated correctly- was the gain on/off, was the tilt down. The radar only shows water droplets not ice, so if it is level or up, nothing shows up at the temperatures one would be experiencing. As I always said to my trainees-: "you have to look at the bottom(of the CB) to understand what the tops might be like".
Given that I have not seen a full report I wont make further comment accept to say especially in the A320/330/340/380 family if in doubt of the warnings, physically sit on your hands until the conflict in your mind is resolved then maybe touch the sidestick.

Whereas it's a great theory for the press and Airbus that 3 experienced aircrew all forgot the basics of how to fly an aircraft simultaneously, it is of course horse manure. This is just another example of Cognitive Dissonance

When the aircraft started departing from the indicated pitch and speed, the crew were forced to use their experience do determine what was really happening. When some things didn't fit that model, the model was likely changed to accommodate that. Everybody mistrusts a computer, but air instruments are supposed to remain reliable. When an air data computer is driven from air instruments and then to the AP, the symptoms of failure now become complex and less intuitive.

Pitch and thrust control will generally keep you alive, but at night, in a storm, with a cacophony of alerts and warnings going off and the AP trying to fix a problem that didn't exist, somehow the standby AI and thrust settings were not recognized as being the only control references that could be safely relied upon.

Any one of these crew could have probably executed a flawless stall recovery if they had recognized what was happening, but at the given flight level it appears the aircraft rapidly got away from them. So once again the history of aviation safety has an example of a relatively benign event - in this case the failure of the pitot heaters - conspiring with other circumstances to build a scenario that proved fatal for everyone involved.

The fact that the pitot failure was a known condition simply adds to the sorry tale of tombstone engineering practices conducted by the OEMs.