Faulty Pitot Tube, Ice & PI* Stall Caused Air France 447 to Crash

50 Comments
Join the Conversation
A-330 Primary Flight Display & Navigation Display - Frank Hardy
A-330 Primary Flight Display & Navigation Display - Frank Hardy
Was advance technology, aviator techniques or a systemic, industry-wide problem at the root of the Air France Flight AF447 accident?

On May 27, 2011, nearly two years after the Paris bound flight crashed into the Atlantic Ocean, French accident investigators released information from the newly recovered "black boxes" that reignited old controversies. Is pilot skill and technique or advanced computer technology the most important aspect incorporated into modern airline design and training?

The Crash

Nearly four hours after departing Rio de Janeiro, Air France flight AF447 (an Airbus A-330-203) plunged 38,000 feet into the Atlantic Ocean killing 228 people. Flight data and cockpit voice recorder information released on May 27, 2011 by the French Accident Investigation Bureau (BEA) indicate seven significant points.

  • The young First Officer (37) and IRO (32)** were at the controls of the aircraft at the time of the accident; however, the captain did enter the cockpit prior to impact.

  • The captain was aware of a potential temperature problem as the aircraft flew through a high altitude cloud layer.

  • There was ice accumulation, caused by super-cooled warm water vapor, rapidly carried to altitude by the updrafts from the thunderstorms the aircraft was overflying.

  • This massive amount of moisture cause the heated pitot tubes to ice over and provide airspeed discrepancies for less than one minute to the co-pilots.

  • The first officers seemingly did not understand the meteorological condition they encountered and incorrectly changed the pitch attitude of the aircraft, which caused it to climb and lose airspeed.

  • Flight 447 stalled and the co-pilots apparently were untrained in recognition and recovery techniques.

  • The airplane plunged 38,000 feet in 210 seconds killing all 228 aboard.

Experience

This crew was, by government and media standards, experienced. According to James Sturcke of UK's The Guardian on June 1, 2009, Captain Marc Dubois had 11,000 total flight hours (1,700 on the Airbus series) and the first officers had 3,000 and 6,600 flight hours respectively (800 and 2,600 on the Airbus series). Nevertheless, the flight data recorder indicated, the cockpit voice recorder orated, BEA affirmed and seasoned pilots recognize that Air France 447 entered what is called a classic deep (or secondary) stall.

The Controversy

This tragedy reignites the long standing controversy over two fundamental principles: technological protection verses individual recognition and recovery.

Technology - advocates believe automation improves safety and the Airbus 330 is one of the most advanced aircraft flying. Investigative journalist and pilot Peter Berendsen wrote in the May 2011 issue of Professional Pilot Magazine, "Airbus engineers believe in cockpit automation, since they have to sell their aircraft to a fast-growing worldwide market with limited supplies of qualified and experienced pilots."

The A-330 is a fly-by-wire aircraft, meaning the pilot does not directly control the aircraft but enters instructions into computers. The CPUs interpret these pilot inputs and manipulate the flight controls as requested. In normal operation (called Normal Law) the computer restricts the pilot from entering unsafe information. However, in cases where computers fail Alternate Law leaves the pilots vulnerable without certain protections. In these situations the pilot can stall the airplane as happened in AF 447.

Technique - advocates believe pilots must always be aviators and until the late 1980s every airline simulator training session began with a skills review of basic airmanship. However, with the advent of advanced computer technology (like the Airbus), there has been a flow away from aeronautical skills toward becoming system managers. Today's pilots are rarely taught the rudimentary aspects of flight, such as steep turns, stall recovery and unusual attitudes, in large airliner aircraft.

Conclusion

Seasoned pilots and investigators believe these fundamental skills can never be engineered and designed. They believe these skills require repetition to become natural and are always necessary for safe flights ultimately controlled by airline pilots. With the advent of new training programs, dictated by aircraft manufacturers like Airbus and approved by government agencies, pilots saw these skills programs disappear. Aviation consultant and former Braniff Airlines pilot John Nance told Lisa Stark of ABC News on May 27, 2011 "...in reality, it comes down to training. What we have here is a major training issue, training and proficiency," Nance argued.

On the PBS program "NOVA: Crash of Flight 447," broadcast several times during 2010 on WGBH TV, another seasoned accident investigator and former US Airways pilot Captain John Cox asked the rhetorical question, "Was this crew one that was...very slow to pick up the unusual [attitude] information,"

Veteran air accident expert Tony Cable, who investigated the Concorde crash (Air France 4590) and Pan Am 103 (the 1988 Lockerbie bombing) criticizes technology on the same PBS program. According to Cable, the computerized airplanes have "raised the question whether the situation is actually being made worse by the increase in automation whereby crews don't get a great deal of opportunity to manually fly the aircraft."

Unfortunately, BEA results indicate that these were the skill sets needed by the young pilots flying AF447. Captain Nance put it best when he argued "I would think one of the major points that has to come out of this [accident] is that pilots -- not just at Air France, but all over the planet -- need to be massively trained."

*PI = pilot induced

**An IRO (International Relief Officer or Cruise Pilot) is another First Officer (or co-pilot)s

Captain Frank , Frank Hardy

Frank W. Hardy - Frank has 36 years of airline experience navigating every ocean & continent. Flying 25,000 hours in 42 years presents a rare historical ...

rss
Advertisement
Leave a comment

NOTE: Because you are not a Suite101 member, your comment will be moderated before it is viewable.
Submit
What is 3+8?
50 Comments

Comments

May 28, 2011 6:00 AM
Guest :
I think it is good.
May 28, 2011 6:45 AM
Guest :
You never pull up on the stick to increase altitude, you push down to increase speed and therefore lift. Pilots nowadays do not have the ability to fly by the seat of their pants, so to speak. They do not understand manual conrol of an aircraft, which is scary, and now proved deadly.
May 28, 2011 7:14 AM
Frank W. Hardy :
Your statement is accurate; however, with the A-330 in normal law the protections will optimize the speed altitude regime and make the nose go down if necessary. The protections, no matter how hard and much you pull up will prevent you from exceeding stall protection. Furthermore, Alpha Floor will automatically increase the thrust to maximum. If this does not prevent the stall the aircraft will automatically lower the nose to the point to maintain flight. So in practice it is possible for the pilot to pull up on the stick to make the nose go down – but it is not natural and will only work in normal law.

With the loss of the speed sensors (due to icing) the aircraft went into alternate law and while it will fly just fine many protections are lost or can be overridden by the pilot. The computer is still flying the plane – just without protections. In direct law the plane flies like the old push/pull planes of yesteryear (B-727, DC-9s etc).

In other words the plane will not push the nose down like in the above example. Sadly, they could have flown this plane if they knew how to recover from a stall.

Frank
May 28, 2011 7:29 AM
Guest :
Terrific article. Succinct and terrifyingly clear. I wondered about studied Gaullic indifference thought. The pilot leaving the controls just before encountering strong turbulence. And the same pilot not immediately taking the controls from a very junior co-pilot. And the controls. Like that ballplayer and his instructor flying the airplane into the building in NYC. The stick on that plane was not analog and could fake out force through springs instead of air force. Could it be, as someone said, a pilot trained in a simulator thought the tilting up of the chair was acceleration instead of climbing? Or did he really not know to go down instead of going up when hearing the stall alarm going off?
May 28, 2011 8:01 AM
Guest :
So (let me see if I understand..) the 'Fly-By-Wire' software has "built-in safe guards" - if you "pull up," in certain conditions, it'll automatically push the nose down, since the "default is to keep flying"; yes?
So in the reported instance, with all the audio/visual warnings going off, and the pilots knowing icing is to blame (they MUST have, seeing that they flew into icing conditions). They'd have shut down those "built-in safeguards", yes?
So once you turned off said "built-in safeguards" you NOW know that "up means up," and to keep flying, you "ease the nose down," no?
Wait a sec -
You mean to say that those Flight Simulators that Air France uses don't train pilots - in certain instances when you loose those "built-in safeguards" - to fly the aircraft, INCLUDING HOW TO RECOVER FROM A STALL..?! (I'm no pilot, but I find this impossible to believe..!! Also, wasn't it reported that just minutes before, other aircraft flew into and safely out of that same storm?)
Me
May 28, 2011 8:30 AM
Guest :
The tale BEA wants to believe us that one inexpert co-pilot steered the aircraft up instead of down against the most elementary anti-stalling rules, and that during several minites, neithet the more experienced other copilot and later the very expert Captain didn't say anyhing.

And that after the disengaging, the co-pilot didn't called back the captain and when he finally did, the latter took leisurly time to comply.

This seems an outrageous plot to disguise extreme grave shortcoming of Airbus as "human errors"of 3 dead persons, by a team of French BEA, Airbus and AF expert analyzing their own nationals misdeeds without neutral supervision!!



The only phrases allegedly recorded were laconic acknowlegment of loss of altitude and speed.!!

If there is anybody that believes that, let himor her explain this!! Even most stoic Scotsmen
would have spoken or yellesd a lot in such emergency!!

The only rational explanation: BEA "forgot" to mention Airbus damaging evidence, or, hopefully not a fact, erased such evidence alleging damage after long time under water!!
May 28, 2011 8:32 AM
Guest :
-----
Aviation experts said it was early and most probably farfetched to blame the miscommands -so basic one compared it to hitting the accelerator instead of the brake when facing a car collision -on a conscious error.

"One of the weird things about this is that the aircraft was definitely stalled, because the crew had had a stall warning, but they were not doing anything to recover from the stall," said Learmount at Flight International.

"It was almost as if they didn't know the aircraft was stalled, because they could have recovered."
-------

Frank, what do you make of the above quote? From http://www.vancouversun.com/mobile/iphone/story.html?id=4856009

Personally I find it hard to believe 3 pitot tubes are sufficient, when accurate air speed measurements are so critical to flight.

Thanks, Krishen
May 28, 2011 8:42 AM
Guest :
"massively trained" -- sounds expensive so it won't happen.
May 28, 2011 9:10 AM
Guest :
A most lucid and correct assessment -- As an experienced pilot friend of mine has just said: The best pilots today are judged by how fast they can type instructions into the FMS!!
Basic flying skills, especially stall recovery, are no longer practised nor assessed in routine airline checks.
May 28, 2011 9:49 AM
Guest :
This is about the third or fourth time that pitot tubes have resulted in this exact incident and exact outcome. How do pilots not have this front of mind at all times? Why have the manufacturers not resolved this issue? Surely there is something that can be engineered that is more reliable than a pitot tube which apparently can be blocked by tape, bugs, ice etc...Makes me think if I'm ever a passenger on a flight that starts to go down I should run up to the cockpit and take over the controls! My flying experience is one glider flight, and stall recovery was the first thing taught.
May 28, 2011 10:12 AM
Frank W. Hardy :
Krishen,

I usually don’t comment on other articles; however, the Vancouver Sun (normally a reliable source) has published a story that should be relegated to the trash heap. It is not only inaccurate in many ways but also uses words that (maybe technically accurate) tend to mislead the uninformed. Furthermore, the wordage is unnecessarily sensational and obtuse. Here are a few examples.

1.“The chief pilot was on a scheduled rest break….” Captain Marc Dubois was not the Chief Pilot but the Captain and aircraft commander of flight 447. The chief pilot is a management position and while they are pilots they are corporate politicians and spokespersons that rarely fly. Every flight has an aircraft commander and some ultra-long haul flights I have flown use multiple captains but only one aircraft commander.

2.“the plane flew into a tropical storm….” Completely wrong and misleading. The actual condition was what is called a squall line and quite small by airline standards. Just ask any pilot that routinely flies through tornado alley this time of the year.

3.“the flight's second pilot tried to rouse the snoozing captain.” Simply an uncalled for, over exaggerated statement. It was 2 o’clock in the morning and the Captain was doing what just about every international captain does 3-4 hours after takeoff. This is the time of the lowest workload and now is the time a captain is supposed to rest to be ready for the landing. The whole purpose for multiple pilots is to allow one to rest while still maintaining a full crew (it only requires two pilots).

4.“According to recorder data, the younger of the two men….” This is a false statement and obviously copied from a German article already proven inaccurate. Nowhere in the document does the French Accident BEA state which of the two pilots were flying (I linked to the actual report.) Sure it will come out eventually but it has not at this time and that statement is unsubstantiated and wrong “recorder”.

5.“the second pilot had begun anxiously calling…after the aircraft's autopilot suddenly cut out….” Misleading! He called because of the speed problems and the fact that the aircraft had degraded into alternate law - not the autopilot.

To go further in this article is worthless, so I say ignore their reference to the three Pitot tubes. Yes they did fail and there was a warning about the icing problem, but they were only a supporting factor in the accident. Similar to saying the tire blew out on a car while driven drunk by a 12 year old down a country road at100 mph trying to elude the police. The tire is only peripheral.

Frank
May 28, 2011 10:24 AM
Guest :
1 December 1974—Northwest Orient Airlines Flight 6231, a Boeing 727, crashed northwest of John F. Kennedy International Airport during climb en route to Buffalo Niagara International Airport because of blockage of the pitot tubes by atmospheric icing.
6 February 1996—Birgenair Flight 301 crashed into the sea shortly after takeoff due to incorrect readings from the airspeed indicator. The suspected cause is a blocked pitot tube (this was never confirmed, as the airplane wreck was not recovered). [7]
2 October 1996—AeroPeru Flight 603 crashed because of blockage of the static ports. The static ports on the left side of the aircraft had been taped over while the aircraft was being waxed and cleaned. After the job was done, the tape was not removed.[8]
February 23, 2008—B-2 bomber crash in Guam caused by moisture on sensors.[
May 28, 2011 10:25 AM
Guest :
Bottom line - These "highly experienced" pilots did not do what every 10 hour pilot in training learns - at the immediate indication of a stall lower the angle of attack by lowering the nose and adding power. -RH FAA CFI AI MEI for over 30 years.
May 28, 2011 10:41 AM
Guest :
Two modern jerks 32 and 37 years old without old school attitude caused that tragedy.There will be many others unless we start retraining these "modern young morons" who know next to nothing.
May 28, 2011 11:29 AM
Guest :
ONE COMMENT STOOD OUT: "QUITE SMALL BY AIRLINE STANDARDS".

2.“the plane flew into a tropical storm….” Completely wrong and misleading. The actual condition was what is called a squall line and quite small by airline standards. Just ask any pilot that routinely flies through tornado alley this time of the year.

"QUITE SMALL BY AIRLINE STANDARDS". ???

Everything I've seen and heard about THUNDERSTORMS in over 40 years of military and airline flying is AVOID AVOID AVOID! And why didn't they deviate? No one can predict the ferocity of a thunderstorm. How did this plane suddenly zoom up to 38,000' if not for extreme up-drafts?
And then with limited control over the plane in Alternate Law and mis-leading airspeed indicators, and lack of "Upset" training, this flight was doomed.

Wally





May 28, 2011 11:39 AM
Frank W. Hardy :
Wally, avoid is the first word and these pilots did deviate to the left of course (about 12 degrees if you look at the DFDR data); however, as a pilot you are familiar with the difference between a tropical storm (one step below a hurricane) and a relatively small squall line. I am not trying to minimize any cell but any seasoned pilot knows the difference between a series of thunderstorms and a tropical storm.

If you are familiar with Atlantic tracks, or any tracks for that matter. Deviation is not simply calling on VHF and zipping off. Sure you don’t fly into cells but deviation in a non-radar environment is not like you may be aware of in radar contact.
May 28, 2011 12:59 PM
Guest :
Frankly it does not matter what speed and other information the pilots were observing or missing, although a hint in the BAE facts' report does seem to highlight that everything turned reasonably normal briefly at 37,500 before more nose up inputs and it seems they were aware they were falling from their discussion about level 100, but to free fall for over 3 minutes and not appreciate they were in a stall and try nose level, would seem to be damning on the crew and will look bad o Air France training. I do not know, but there ought to be a procedure they should have but did not follow. This may turn out to be worse that Aeroflot A310 debacle.
May 28, 2011 3:10 PM
Guest :
Mr Hardy,
An excellent piece on this tragedy. As a former military pilot, it pains me to see pilots seemingly forgetting their basic airmanship skills. They were obviously confronted with a substantial number of problems and seemed to become overwhelmed by the situation-much like someone swho becomes spatially disoriented and will not believe their instruments. It certainly appears they had no reliable airspeed information and since I am not an Airbus pilot, I don't know how much instrumentation they lost, but to ignore stall warnings and not "feel" how their airplane is flying is too much for me to take. Aviate, navigate, communicate-it is time for everyone to remember and get back to the basics. I can only hope the loss of this crew and all these passengers will provide a wake-up call to the industry. This was such a senseless loss of precious lives.
Kevin
PS-hope I don't have too many misspellings as my machine will not let me scroll back up and review.
May 28, 2011 3:10 PM
Guest :
A pilot with good airmanship can always fly a plane with basic instruments, controls and engines functioning. In Air Canada Flight 143 ("the Gimli Glider") the pilot was able to land a 767 with no engines. Captain Bob Pearson was a glider pilot and this would have honed his basic airmanship skills. As you say, Airbus Captains are mangers of computers rather than aviators. After all, it is much easier to enter your requirements into a computer and let the plane fly itself rather than take the controls manually. The problem is, unless you are flying the plane manually on a regular basis, it would be difficult to immediately come on line when required. I would imagine that the pilots in Air France 447 spend much of their time using the A330's computers to fly the plane and little time actually handling the controls themselves. How could three allegedly seasoned pilots fail to notice an altimeter which showed rapid loss in altitude, attitude indicator showing nose up and a throttle that widely open (with engines functioning) and not act instinctively. If these pilots were aviators they would have immediately put the nose down. A lay person would have thought, the plane is losing altitude, I have to put the nose up! For similar reasons, Colgan Air Flight 3407 stalled and did not recover.
May 28, 2011 5:35 PM
Guest :
So basically - faulty pitot tube: Incorrect airspeed indications: pitch up causing the aircraft to enter a deep stall, and crash into the Atlantic...

The lives of 228 people were at risk, and pilots these days have become far too reliant on this fly-by-wire system in the Airbus planes, correctly demonstrated by the complete lack of control and training in this unfortunate accident. They had receieved a stall warning did they not? I'm certainly no pilot, but pitching downwards will enable you to GAIN airspeed and therefore recover from an imminent stall? Training needs to be hitched up a notch. This is no game. The transportation of lives 35,000 feet in the air is not something to be taken lightly.
Pitot tubes need to be maintained, or the misreadings will cause a catastrophic chain of events that will result in the deaths of many people. It's sad to know that prior training could have enabled these young pilots to avoid a potential dangerous situation. Pitot tubes have long been causing problems, and something needs to be done to prevent them from icing over. Good informative article.
May 28, 2011 6:30 PM
Frank W. Hardy :
The comment “Level 100” simply means to me the PNF was aware they were passing 10,000 feet on standard QNH. This is quite normal, for some airlines make not only the “one thousand to go” or “one thousand above/below” or “nine for ten” calls but also a call at every 10,000 feet i.e.: [Passing] “Level 300” or [Crossing] “Level 100.” I have been unable to verify whether this is a standard call at Air France for my friends at AF/KLM have not returned my emails for obvious reasons.

As for Captain Bob Pearson and Air Canada 143 – well Captain Pearson had a lot of things going for him – a whole lot of luck! Not only was he a glider pilot but also his first officer Maurice Quintal was a former RCAF pilot who knew about Gimli airpark (a former RCAF air base). Had Maurice not been in the Canadian Air Force, had he not known about the field or not remembered exactly where Gimli was, had Bob had another copilot that July day, Captain Pearson’s glider experience and excellent airmanship, while good, would not have saved the day and probably 69 people would be dead!

There are other similar problems, Air Transat 236 over the North Atlantic. Another glider but this time an Airbus 330. Had these guys not been rerouted on a more southerly track they would never have made it to Lajes AFB.

What about the 155 people Sully Sullenberger, the Hero on the Hudson, saved in his Airbus glider? Sully too is a glider pilot but I believe he would argue Jeff Skiles, the location and conditions on the Hudson were extremely necessary that day.

I do not wish to minimize the requirement for skill, for that is the fundamentals of this article; however, skill is only a vital addition. Pilots need luck too like the guys above. There are unlucky pilots with lots of skill.

Example, Ethiopian 961 and the 175 folks on that B-767. This flight encountered all the situations confronted by the guys above. Captain Leul Abate had the same problem as Captain Bob Pearson with a fuel starved Boeing 767. Abate confronted the same open ocean glider situation as Captain Robert Piche of Air Transat 236. And Ethiopian 961’s captain faced the same ditching situation as Captain Sullenberger – but Captain Abate was unlucky.

There was no airfield within gliding distance. There was no military base situated in the middle of the ocean and the Indian Ocean was rough that day with high seas and not calm like New York’s Hudson River was for Sully – Captain Leul Abate was skilled but unlucky. His skill saved the lives of 50 people but unfortunately luck killed another 125.

Whether it was Jeff or Maurice, a reroute or abandon airfield, pilots need both excellent skill and a lot of luck.
May 28, 2011 8:12 PM
Guest :
Why is there an argument about this? The data recorder and CVR clearly indicated that the stall warning had activated and there was no response from the pilots to reduce the angle of attack of the aircraft when it did. Bad airspeed indications, bad weather, Alternative Law - all not important. Every pilot knows or should know that the stall warning system is independent of the airspeed system and is accurate at any airspeed or flight attitude even when the airspeed indicator might not be accurate. Immediate shoving of the nose down upon hearing the stall warning would have saved this plane and all aboard.
May 28, 2011 10:59 PM
Guest :
I am an Airbus A330. My company regularly conducts stall and jet upset recovery during our sim recurrency.
May 29, 2011 12:48 AM
Frank W. Hardy :
For the Airbus pilots who have contacted me regarding training, I do agree that there have been a number of airlines that started retraining programs in high altitude stalls and upset conditions. But let me ask one simple question, when did they start that rudimentary training? This accident happened two years ago (4 sim sessions).

Yes there are also a few airlines that never completely ended the program but only reduced the number of maneuvers and or the required level of proficiency. Some airlines do this as a demonstration event – “OK here is what XXXX looks like.”

United and US Airways are two of the most prominent airlines that continue to train pilots in not only recovery but upset and crossover affects; however, they both had nasty upset accidents (UAL 585 in Colorado Springs and US Air 427 in Pittsburgh).

When we discuss my airline does this or that airline does that, it is not necessarily an industry practice and that is what must change. Any of you pilots that use FCOMs for the Airbus (320, 330 or 340), please find those maneuvers in the SOPs!

We must remember unreliable airspeed is now a memory item and when I started flying in the 1960s the old timers use to tell me – “pitch and power son, that’s all you need to remember” and we constantly trained to that. An ILS approach (for example) is flown by fine tuning a VSI and fuel flow based upon wind and weight - pitch and power.

Well then we got away from that and went to gee whiz airplanes, dual channel stall warning systems, GPWS (and later EGPWS) and a whole bunch of other nice things. All that produced was Birgenair 301, AeroPeru 603, Flash Air 604, Gulf Air 072 and a lot more. But pitch and power could have saved those flights.

So today (and for the last several years) because of what we learned from those accidents just about everyone trains for unreliable airspeed (and spacial disorientation). What is it 15 degrees/TOGA; 10 degrees/TOGA; 5 degrees/Climb – THAT’S PITCH AND POWER FOLKS! Exactly what those old time Captains told me when they were flying DC-6s, Super Connies and Convair 440s. As Yogi would say, “It’s de ja vu all over again!”
May 29, 2011 9:48 AM
Guest :
Thanks for an informative article. When a plane loses lift going too slow or going too fast, wouldn't you as a pilot be more afraid of the latter, at least if you're high in the sky? With the A330 descending rapidly, would the pilots know for sure it was going too slow when speed sensors didn't work?
May 29, 2011 10:46 AM
Guest :
Thanks for an informative article. When a plane loses lift going too slow or going too fast, wouldn't you as a pilot be more afraid of the latter, at least if you're high in the sky? With the A330 descending rapidly, would the pilots know for sure it was going too slow when speed sensors didn't work?
May 29, 2011 11:18 AM
Guest :
why are GPS data not used as back up to the exisitng speed instrusment.
May 29, 2011 12:20 PM
Frank W. Hardy :
In a modern turbojet a pilot cannot tell how fast he or she is going at 35,000 feet by looking out the window. You cannot tell (without a visual reference) within several hundred miles per hour so no it is not possible to determine if you are in a high speed or low speed stall without instrumentation and it really does not make much of a difference. The recovery is different but the result is the same without proper corrective action. And because vertigo is always possible, seat of the pants flying is not recommended at all.

As far as too fast or slow - well you can tear the wings off by going too fast or tear the wings off by going too slow. It really doesn't matter since planes can't fly without wings.

As for GPS, you are absolutely correct. It is used and a required memory item for unreliable airspeed in a bus; however, it boils down to recognition – what do you have? We all have the impeccable vision of 20/20 hindsight but in the heat of battle you have to draw on standard procedures. If that fails you drawn on previous knowledge! If that fails you draw on years of airplane experience and if that fails you draw on all that you know about flying. The further down that list you can go the greater your chances of survival become.

How do you tell a new EFIS pilot?
He asks, “What is it doing?”

How do you tell a junior EFIS pilot?
He says, “Look at what it is doing.”

How do you tell an experienced EFIS pilot?
He says, “Look, it is doing it again.”

“There are old pilots and then there are bold pilots; but there are no old, bold pilots!”
May 29, 2011 1:26 PM
Guest :
Thanks for an informative article. When a plane loses lift going too slow or going too fast, wouldn't you as a pilot be more afraid of the latter, at least if you're high in the sky? With the A330 descending rapidly, would the pilots know for sure it was going too slow when speed sensors didn't work?
May 29, 2011 2:50 PM
Guest :
Very illustrative article. This recalls me the crash of VARIG RG 254 in the Amazon forest on september 3, 1989, due to fuel starvation. Instead of inserting a heading of 27 degrees, the pilot inserted 270 degrees and at no moment he could realize that he was flying west, instead of north, although he had the sun ahead all the time (it was in the afternoon).
May 29, 2011 4:21 PM
Guest :
Would the corrective actions required by the pilot be the same regardless if the plane flew too fast or too slow, and would a wrong guess be equally catastrophic?
May 30, 2011 5:16 AM
Frank W. Hardy :
There have been many fuel starvation problems over the years but the saddest was Avianca 052 that crash near KJFK. Lack of English by the Captain, improper interpretation (or relay) of the Captain’s desires by the First Officer, a wind shear on final and a Kobayashi Maru senior at the end. When encountering the wind shear on short final if you exercise the wind shear recovery action you will run out of fuel. If you do not exercise the recovery action you will crash due to wind shear – a true no win situation for those guys.

One of the good things that came out of that accident was the requirement for all pilots on an international flight deck to have a better understanding of English.

As for the recovery actions of an over-speed verses under-speed stall; the actions are as you would think – opposite. One adds power and push and the other remove power and pull (Basically). This is for a positive dihedral western style aircraft that have pitch up power tendencies. The procedure is opposite for a series of Soviet style airplanes with negative dihedral or high wing/engine combinations mainly found in certain multiengine land or seaplanes.

And Yes, doing the wrong procedure would be very bad - think stepping on the accelerator and flooring it when you are approaching a group of stopped car at an intersection.
May 30, 2011 6:26 AM
Guest :
We all know that stall recovery procedure for all aircraft from piston powered to the latest jet transports made by absolutely every single aircraft manufacturer in the entire world is to lower the nose and add power. Every student pilot is taught this early in his or her training. But the recently recovered flight data boxes indicated opposite inputs in the case the ill fated Air France flight 447. Since all pilots know this proper procedure why did these three (3) fully qualified experienced pilots NOT apply forward nose down inputs? I have to ask, is it possible that this computer controlled machine would not allow human input?
Recent information from Paris = "Air France and Airbus are facing manslaughter charges following the crash and a judicial investigation led by Paris judges is under way." Dose someone in France other than the official government investigators have a different theory than the presently suggested "pilot error" as the cause? Since there could be a fundamental government desire to protect the government owned Airbus company from having to take a serious look at the Airbus design belief in extensive computer control over all modes for flight, is there a need for a truly independent investigation rather than one that has a clear connection to protecting Airbus? After all, if it turns out that the multiple computers on all Airbus airliners and not the pitot tubes were the cause of this disaster then all Airbus transports are not just suspect but unworthy of flight. The French government cannot think kindly of such a conclusion. And since the German government has an equal liability it the Airbus venture, this is a serious European problem. Indeed, to a smaller extent so does the British government as well as others.
Since no less than thirty-two (32) incidents of spurious computer inputs on various Airbus airline flights have been reported within the US over the last few years we have a serious American interest in this investigation. These events ranged from minor turn commands to serious flight deviations. Surely these undesired actions along with the fatal flight 447 in an Airbus 330 would be enough to include our NTSB (National Transportation Safety Board) in a parallel investigation.
Once again I am reminded of the movie "2001" staring the rouge computer, "HAL" that got carried away and very nearly killed the entire mission. Also there is that old joke, "Ladies and gentlemen, sit back and relax. This aircraft has no human pilots and is completely computer controlled. So, nothing can go wrong, go wrong, go wrong, go wrong…"
My credentials are: Naval Aviator, thirty years with Pan American, four years with United and last ten years as captain 747 with a total time of 23,000 hours. What say you?
May 30, 2011 10:11 AM
Frank W. Hardy :
I say you sound a little biased against Airbus and Europe. Let us not forget the US FAA would not ground the 737 after it was believed (and later discovered) that the 737 PCU/rudder assemblies were faulty. We know that there was a design fault and some people believed it after the USAF had a T-43 crashed during navigator training – they blamed that on the pilots. By the time UAL 585 crashed it was all but certain, but still no action and this time they refused to make a determination and let the plane keep flying. Then we had US Air 427 and another 133 lives lost and still no US action.

The government was quite straightforward. They remember how the 1979 shut down of the DC-10 nearly bankrupt Continental and said grounding the 737 would put Southwest out of business. A short time later we nearly lost a whole bunch of people in Eastwind Airlines Flight 517 – same problem. So let’s not be so harsh on our European friends.

Heck the USA can pull the Type Certificate on any airbus they want at any time. Remember the FAA certified the procedures and training for US airlines, not France. And just like after 585 old pilots knew something was wrong, this old pilot is saying it is not the plane but the program.

I’ve been a check airman in the US and a TRI in Europe and Asia – I’ve seen some really good pilots do some really bonehead things in the sim. That is a place where they are ready for it so I am not so quick to point blame here. I do criticize the training the industry has undertaken these last years but not the pilots – I was not in that cockpit.

A lot has been made of the Captain. We must remember at an IVSI of nearly 11K FPM it is highly unlikely he could get into the seat even if the F/O could get out.

Now your comment about stall recovery, well I do not completely agree with you. While the pilot is to lower the nose you are not to lose altitude in the process. The link below will take you to a photo of an Antonov AN-72. If you will notice adding power will not only push the nose DOWN but guarantee you will lose altitude. I’ve talked with a ton of AN-72/74 drivers and their stall procedure is different from what you are used to.
http://www.airliners.net/photo/805114/
May 30, 2011 11:03 AM
Guest :
I have read the translated version of the BEA release from yesterday and many articles including yours. Why does no one discuss the change in thrust lever position from TO/GA at time 2h 10min 51 to IDLE at 2h 12min 02? Someone changed those positions and that would indicate a significant reduction in thrust. Is this not something worth analyzing?
>
May 30, 2011 11:09 AM
Guest :
No way. There is no way that they weren't trained in stall recovery. Clearly, amidst a very confusing cockpit environment (conflicting speed and altitude information, possibly several different warning alarms), they lost their situational awareness, got away from the critical "fly the airplane" mentality, and allowed what should have been a temporary problem to compound into a terrible tragedy. To say that pilots with 3000 and 8000 hours didn't receive training in fundamental manoeuvres like stall recovery is way too presumptuous. Notice how A330s don't crash every day - it was a rare freak accident. Also, it is incorrect to place blame entirely on 1 factor, as any time an accident like this occurs, it occurs as the result of a chain of events. If nothing else, whoever designed the pitot tubes on the A330 is culpable, because there would have been no incident at all had their pitot tubes not frozen over. Please stop saying that they didn't know how to recover from a stall, its way more complicated than that.
May 30, 2011 11:13 AM
Guest :
If the pilots on the A330 were convinced the plane was going too fast, maybe they never tried full throttle. However, would it even be possible for a plane to go too fast in a descent when motors have been idling for minutes?
May 30, 2011 12:05 PM
Guest :
Massive computer systems failure?

John T Halliday wrote an article in the Huffington Post in August 2009 describing a possible scenario. I am not an aviation expert; does this seem plausible to the professional pilots posting here?
May 30, 2011 12:06 PM
Guest :
Here is the article I mentioned:
http://www.huffingtonpost.com/john-t-halliday/air-france-447-a-cockpit_b_ 268672.html
May 30, 2011 2:24 PM
Guest :
Did the pilots think that the artificial horizon and altimeter were also not functioning? What is the relevance of statement about loss of visual reference and why would this be necessary? Was there any evidence from the CVR that the diagnosis was overspeed? Should these pilots not have relied on the available instruments and available thrust to make decisions on the correct course of action?
May 30, 2011 2:43 PM
Guest :
If the pilots on the A330 were convinced the plane was going too fast, maybe they never tried full throttle. However, would it even be possible for a plane to go too fast in a descent when motors have been idling for minutes?
May 30, 2011 10:32 PM
Frank W. Hardy :
Visual reference at night over the ocean is almost impossible. Spacial disorientation is very easy (think JFK Jr.) but having said that in a modern airliner looking outside the plane is something you do in three situations. To look for lightning to give you an idea about the cell you see on your radar, to look for traffic you get that comes up as a TA on your TCAS and most often to look at the wonder that few humans see – mother earth from our perspective. Other than that you rely on your instrumentation.

On the A-330 (and other models of the Airbus) the thrust levers (throttles, power handles, motor sticks, go-poles etc.) do not move. The only way to tell if the engines are at idle or at TOGA power (full power) is to look at your ECAM, Electronic Centralized Aircraft Monitoring System, (EICAS for you Boeing drivers) and physically look to see what you have. In an upset situation it is extremely easy to think you are over speeding when in fact you have an under speed stall. But the airplane’s notification system is different for each and it is possible that there was an over speed warning for a few seconds during the initial unrealiable airspeed situation.

That condition could be sufficient to have the pilots disbelieve either system since every pilot knows it is impossible to have an over speed and under speed stall simultaneously. Yet, as we saw in the crash of AeroPeru 603, the warnings do happen at the same time. Nothing in the BEA report indicates it did not happen and we must LISTEN to the CVR to hear if it did occur for a few seconds initially.

The vertical acceleration could have further disoriented the young pilots. Remember this plane reached nearly 11,000 fpm descent rate. For an A-330 that is astonishing. Having a few thousand hours in the plane myself (both A-330-200 & A-330-300s) it is very hard to reach that sink rate even if you tried with autopilot disconnected, full speed brakes extended and accelerating toward MMO. I think the greatest I ever reached in a simulated High Dive (rapid decompression) was about 7000-8,000 fpm. So since they were going this rate it is highly possibly they thought they were over speeding when in fact they were doing slightly over 100 knots (deep stall speed).

Hope this sheds some light on your questions.
May 31, 2011 4:40 AM
Guest :
The takeaway from this is that instead of the autopilot disengaging when the pitot tubes malfunction, the autopilot could automatically put the plane in the pitot-tube-failure configuration of nose up 5 degrees, thrusters at 80% (according to PBS' "Crash of Flight 447" at http://video.pbs.org/video/1685933496). The engines were still good. The plane kept upright (even under stall conditions). They just needed to put the nose down to get out of the stall. The pilots were overloaded with information and their sudden dramatic increase in workload. The computer would have had no such overload.
May 31, 2011 8:17 AM
Guest :
The AeroPeru 603 crash appears to have many of the same elements discussed here, though ATC was available in Peru. If nose up and idle engines is explainable behavior in an over-speed stall, isn't a possible explanation that the pilots (and/or the computer) on AF447 never realized the speed was in fact too low?

Still, if the pilots were in control, why didn't they try the opposite action eventually? Rational thinking cannot be taken for granted, but just doing the same sounds like a rabbit caught in the headlights.
May 31, 2011 11:43 PM
Guest :
Would using the thrust reversers possibly tip it out of stall
Jun 1, 2011 2:05 PM
Guest :
On a moonless night why would the pilots even bother to mention that there is no visual reference? Did they think that the artificial horizon was inoperative? The CVR should have indicated what they were thinking at the time. The nose up of 15 degrees is crucial to the crash which they could have diagnosed from the artificial horizon.
Jun 1, 2011 3:51 PM
Guest :
It would be in the interest of the investigators (the French BEA) to protect Airbus and Air France in this tragedy. These are multimillion dollar industries and both companies are major contributors to the French economy. There is a conflict of interest. The deceased pilots would make an ideal scapegoat and can be blamed to protect Airbus and Air France. What is needed is an impartial and unbiassed assessment of the accident. What say you?
Jun 2, 2011 3:15 AM
Frank W. Hardy :
I would agree, an unbiased assessment of the accident is what is needed, but unfortunately that does not always occur – especially now since everything is so money oriented. Sometimes you have really dedicated folks who doggedly study things, looking for the truth and recommending a solution to save lives. Captain John Cox comes to mind and his work on US Air 427. Everyone, the industry, Boeing, GE, Parker Hannifin (manufacturer of the rudder PCU assembly), the FAA and even the NTSB (initially) wanted to blame anything but the plane. Pilot error, wrong inputs, wake turbulence, mountain rotor (UAL 585), disgruntled crew, Granny’s false teeth and just about anything you can imagine – everything except the true culprit. I fully believe we would not know today if it had not been for the likes of John Cox (who I know personally.)

Unfortunately, even John could not operate in today’s environment and I do not think we will ever get a true picture. They operate under a philosophy of I cannot be wrong so that cannot be the problem. Loss of life is the cost of doing business (like breakage in a store selling glass). Of course they do not admit that but it is their modus operandi.

The truth can be expensive and to paraphrase Jack Nicholson in a Few Good Men. The industry tells the people what the people need and how they need it. The people want cheap fares – not the truth. They can’t afford the truth.
Jun 3, 2011 12:49 PM
Guest :
As a pilot, even if you didn't contact ground about problems, wouldn't you word them? Like "I don't know if we're going too fast or too slow" or "the computer blocks me" etc. Aren't you trained to stand on each other's shoulders and pass on as much info as possible?
Jun 21, 2011 6:02 PM
Guest :
This is a good article and a better dialogue. I am just an airline passenger who needs to know what happened so I can rest/fly easy. My question is this, "There must be some previous experience with the same stall, in the same meteorology, the same ice, same equipment, but that was overcome by pilot correction. Am I wrong? I would really like to hear this stall recovery testimony.
50 Comments
Advertisement

Related Topics

Advertisement