Pilot Error Likely Played Major Role in Air France 447 Accident

Accidents/Incidents, Air France

When the French found the black boxes from Air France flight 447 nearly two years after the A330 airplane crashed in the Atlantic off Brazil, it was an incredible feat. But now, the French probably are wishing those black boxes remained on the floor of the ocean, because its national airline is about to face some tough questions regarding the actions of its pilots on that flight. No airplane accident happens because of just one problem, and this is no exception, but so far pilot error is really sticking out as the single largest contributor here.

Un Airbus A330 d'Air France
Photo of Sister Ship to Crashed Airplane via Flickr user Tab59|CC 2.0

The French accident investigation group, the BEA, has put out an update on its investigation around what caused Air France 447 to crash (pdf) in the Atlantic back in 2009. Flightglobal has a good minute-by-minute breakdown of what all of the technical verbiage means, but let’s focus here on a few key points.

Pilots Were Not Inexperienced
One thing that has been picked up on elsewhere is that the Captain was not in the cockpit when this all started happening. That’s true, and it’s not a surprise. That’s why there are three pilots on longer flights like these. They rotate taking rest and this was the Captain’s turn. Does that mean that there were two inexperienced fools manning the controls? No. The co-pilots were highly trained and should have been able to handle this situation without needing the Captain. As Flight notes, one of the co-pilots had more time on the A330 than the Captain himself (just not in command). Experience shouldn’t have been the issue.

Turbulence Was Not a Factor
The aircraft went down in an area near strong equatorial storms, so many people assumed that the storms and the likely associated turbulence played a role. That no longer appears to be the case. The pilots were actively working their way around the storms, and while there was turbulence around, it doesn’t appear to have been anything severe. The storm likely did play a role in that it caused the pitot tubes to freeze over. Let’s talk about that . . .

Frozen Pitot Tubes Are the Likely Trigger
I don’t believe this has been officially confirmed, but the belief remains that the pitot tubes froze and that kicked off the problems on the airplane. Pitot tubes are little pokey-looking things that stick off the side of the airplane and measure airspeed. If the pitot tubes froze as expected, then speed readings would have been erratic and incorrect. That would have caused the airplane to shut off the autopilot as happened here. While it is a serious issue, it shouldn’t have cause and accident on its own.

Ultimately, the Pilots Screwed Up
Regardless of what happened with the pitot tubes, what happened next seems just unbelievable and certainly casts a great deal of blame on the pilots even though we won’t have the final report until next year. About 10 minutes before the autopilot shut off, the pilots noted that they couldn’t climb any higher than the 35,000 feet they were at because of their weight and the relatively warm air outside. In other words, if they climbed higher, they wouldn’t be able to generate enough lift. That makes what happens next even more strange.

When the autopilot shut off, the pilots should have worked to keep the plane flying as it was. After all, there wasn’t an actual speed problem but just a speed measurement issue. The engines worked just fine, so it should have been quite possible to keep the airplane on its path. That’s not what happened. Over the next four minutes, the pilots pulled the airplane into a climb and right into a stall and that led to the crash into the ocean. This goes against one of the most basic rules of flight.

If Your Airplane Stalls

When an airplane stalls, that means its angle of attack (the angle of the wing as compared to the direction of the air) is too great. Fixing it is pretty straightforward and it’s something that gets trained at very basic levels. As the FAA says in its Airplane Flying Handbook:

Reducing the angle of attack is the only way of recovering from a stall regardless of the amount of power used.

That means pushing the airplane’s nose down until the air once again runs smoothly over the wings. If you’re at 35,000 feet, don’t worry about losing altitude. Just get the airplane back into normal flight. How do you know if you’re in a stall? This is where the Boeing vs Airbus people will start their “mine is better than yours” fight.

On Boeing airplanes, the control column actually shakes to warn the pilot. (It’s known, unsurprisingly, as a stick shaker.) But most Airbus types, including the A330 that crashed here, operate with little joysticks on the side and these don’t have stick shakers. Instead, there is a very loud verbal warning repeated multiple times. Either way, it shouldn’t be missed. But don’t Airbus airplanes have greater automation to prevent these things anyway? Not in this case.

Airbus normally has automation protection that prevents pilots from doing something stupid like going into a steep climb in a situation like this, but those protections weren’t in effect because of the inaccurate airspeed readings. That pushed the airplane into Alternate Law which shuts down many of the protections that are in place during Normal Law.

When the Captain got back into the cockpit, the airplane had an angle of attack at an incredibly high 40 degrees and it was losing 10,000 feet per minute in altitude. Despite his best efforts to recover, it was a failed effort. The airplane hit the water with its nose up 16 degrees but still losing more than 10,000 feet per minute in altitude. I can’t imagine how awful those few minutes were for the passengers.

But the Pilots Aren’t To Be Blamed Completely
The final report hasn’t been issued and won’t be until next year, but it’s easy to see from this that the pilots and the pitot tubes were the two biggest contributors. Why did the pilots continue to apply nose-up pressure when that was the exact opposite of what would have happened? We’ll never know what was running through their heads, but it’s easy to see that they could have been distracted.

Remember, the pilots were already working to pick their way through the worst of the storms. Add to that the loss of the autopilot, dozens of failure messages, and inconsistent speed readings and it seems like the answer might be simple. The pilots may have been so distracted that they forgot to do the one thing they needed to do to survive: fly the airplane. Once the final report is issued, look for training changes to come out of this and possibly even some changes in the way Airbus puts its airplane logic together.

Get Cranky in Your Inbox!

The airline industry moves fast. Sign up and get every Cranky post in your inbox for free.

68 comments on “Pilot Error Likely Played Major Role in Air France 447 Accident

  1. Brett, I respect your knowledge of the airline industry, but I think your are operating outside of your envelope. I am not going to defend the pilots, I have a few questions about their performance but “The final report hasn’t been issued”. Its not just about the STALL, but also Mach buffet. You can lose control from going too fast at altitude. Funny thing is when a pitot tube ices over, the airspeed indicator acts like a barometer, ie; you go up the airspeed increases, go down it decreases. This may explain the nose up attitude. Perhaps they were trying to control an incorrect airspeed indication. Years ago Northwest lost a 727 over upstate NY for this same reason. You mentioned that “Turbulence Was Not a Factor”, but did you see the plot of 447 over the GOES satellite photo of the weather at the time of the accident? The other aircraft did serious deviations to get around the system of weather, while the crew of the 447 drove through the thick of the weather. The BBC did an excellent documentary on the accident using the available data, it appeared on NOVA on PBS, look it up! Why didn’t they just push the nose down? In some stall situations you can get a blanking of the tail surfaces that block the airflow (this is why you see parachutes on the tail of test aircraft) and prevent recovery. Also, adding power in under wing engines causes pitching up, just what you don’t need in stall recovery. I will again repeat that I am not excusing the pilots, just saying that the STALL was the end game. What happened before the STALL is the cause of the accident.

    1. Funny thing is when a pitot tube ices over, the airspeed indicator acts like a barometer, ie; you go up the airspeed increases, go down it decreases. This may explain the nose up attitude.

      The report shows an immediate nose-up input from the pilot upon disconnect of the autopilot. It then shows a drop in airspeed from 275kt to 60kt. So the nose-up input did nothing to increase the displayed airspeed. In fact, later, the displayed airspeed did increase after the one time the pilot did make a nose-down input and the rate of climb slowed.

      You mentioned that “Turbulence Was Not a Factor”, but did you see the plot of 447 over the GOES satellite photo of the weather at the time of the accident?

      There’s no question that there were storms in the immediate area, but there is nothing to show that the airplane encountered anything significant in terms of turbulence. There was a warning that they would be entering air that would “move about a bit more than at the moment” from the pilot but I have seen nothing to indicate that there was severe turbulence. There’s only a mention that the level of turbulence increased slightly a couple minutes before the autopilot disconnect.

      Why didn’t they just push the nose down? In some stall situations you can get a blanking of the tail surfaces that block the airflow (this is why you see parachutes on the tail of test aircraft) and prevent recovery.

      That could explain why the nose wouldn’t go down, but it wouldn’t explain why the pilots would try to put the nose down. This report shows the inputs from the pilot, not just the actual movement of the airplane. Had the pilots attempted to go nose down but failed, then that would point toward an issue with the airplane. But the pilot immediately tried to go nose-up.

  2. Agreed with Donald .. speculation and media drama will really abound when those who don’t actually fly the 330 begin assuming what errors led to this tragedy (and let’s also stay away from trashing this airline too – their record and that of the Airbus is pretty impressive).

    1. I did not trash the airline at all. I only mentioned Air France at the very top and said that it would face some tough questions about the actions of its pilots. I don’t see how you can refute that, because that happens in any accident where pilot error is involved. I know nothing of Air France’s training procedures and I have no reason to assume they’re deficient in any way. My mention of training and aircraft changes at the end was not Air France-specific.

    2. the pilots were at litle fault in this crash. you mention that they incorectly responded to a stall however the stall warning ocured shortly after the airspeed dropped to 60kts which may have confused the pilots and assumed that the stall warning was due to unreliable airspeed, this he was aware of. however he wasnt trained in high altitude recovery situations or even flying high altitude without autopilot which is and was common. as the flight was at night over water he had litle spacial awarenes and therefore may not have realised untill it was too late that he was climbing.

  3. Is it time for computers to start giving advice? Should there now be an audio voice saying point the nose down if the computer see they are about to stall or whatever the the problem may be?

    1. Airbus actually has these types of protections built-in when the aircraft is operating under Normal Law. But when things go south, the airplane gives more latitude to the pilots to try to recover. It also helps prevent the airplane from doing something stupid because it may have faulty data.

    1. When flying a plane, what is important is your speed in relation to the air around the plane (i.e. accounting for the effects of a headwind/tailwind), not the speed in relation to the ground. The pitot-static system will tell you this, a GPS cannot.

  4. I have to believe that this incident is more complicated than the pilots forgetting to fly the plane. It seems absolutely inconceivable to me that the pilots could be unaware of the nose-up angle of the plane for the entire duration of the incident…I just can’t believe they could allow the plane to drop to the ocean with a nose-up angle of 40 degrees for three or four minutes without either knowing about it, or trying to correct it.

    1. It seems absolutely inconceivable to me that the pilots could be unaware of the nose-up angle of the plane for the entire duration of the incident

      They were most definitely aware because they were causing it. At least, they were making inputs for most of those few minutes to put the aircraft in a nose-up position.

  5. “Is it time for computers to start giving advice?”
    As Brett mentioned the computers (There are more than one commercial aircraft) did give advice, ie; audio alert of impending stall. There are many other warnings too. I would add that many of the electrons fled to the tail and hid. What I mean is that when the airspeed indicators failed to provide accurate data, the autopilot and auto throttles clicked off and reverted to a lower level of operation that required pilot input. Airliners give many warning of the stall, as airspeed approaches stall speed a red warning indication occurs, then a sticker shaker activates and finally the stick pusher pushes the yoke forward (This is in aircraft with a control column and yoke such as a Boeing). I am not familiar with the Airbus. Unfortunately the computers are getting their information from the same place the pilots are.

  6. @Turbulence Forecast: GPS will tell you the ground speed which isn’t relevant. What is needed is the air speed – how fast the plane is moving through the air. The air itself is moving (often fairly quickly) which is why the plane’s air speed and the ground speed diverge.

    It is unlikely that all 3 pilots had no basic airmanship skills which is why waiting for the final report is better than armchair quarterbacking. Airbus and Boeing have already adjusted training for dealing with stalls and similar situations involving dodgy airspeed information.

    As with the vast majority of plane accidents, they are extremely rare, have multiple contributing factors and every effort is made to learn from them. If only the same thing happened with cars – a 911’s worth of people die every month on the roads.

    1. It is unlikely that all 3 pilots had no basic airmanship skills which is why waiting for the final report is better than armchair quarterbacking. Airbus and Boeing have already adjusted training for dealing with stalls and similar situations involving dodgy airspeed information.

      I’m not suggesting that they had no basic airmanship skills. These were highly trained pilots. It’s just possible that the basics escaped them when they were battling with a variety of different issues. It should be noted that only when the captain returned to the cockpit were consistent nose-down inputs given, so lumping the actions of all three into one statement probably doesn’t do the situation justice.

  7. GPS tells your ground speed, not airspeed and have no bearing on the stall. When the flight attendant would forward the question from a passenger “How fast are we going”? My reply was, do they want our indicated airspeed, true airspeed, mach number or ground speed? This thread is going to consume me

  8. Have you flown in bed weather IFR and have you had all instruments go out on you or show inconsistent readings? One speed shows below stall speed, other speed indicator shows above allowed speed. No horizon ahead of you, you are in a cloud. No stars above you, no city below you. Up, down, left or right, upside down or right side up. Your body has no idea.
    When Kennedy flew his plane into the ocean at night due to his disorientation, he could turn his autopilot on and possible save himself because all instrument were reading correctly and feeding good data to the autopilot. At this time, even the autopilot was so confused from the inconsistent readings that it gave up.
    Have you looked into a A330 cockpit? The only analog display are the artificial horizon (captain side), and Rate of Climb/Descend indicator (VSI) (co-pilot side). Everything else is digital.
    The artificial horizon had to operate as it was running on a mechanical gyro, not computer. If the nose was so up, as the report states, and stall horn went off multiple times, still, the pilot kept the angle of attack 35+ sometimes 40%. Why did the pilot kept the nose up for the next 3 minutes? The analog rate/descent indicator (VSI) tried to show minus 10,000 feet/Min. , but it can’t. it shows 500,1000,1500,2000,2500, and some may shows 3000. I don’t know where the needle point to on the VSI at positive or negative 10,000 feet/min. I never tried it.
    It all points to pilot error. But for 3 and half minutes? Two senior pilots? I keep coming back to – no instruments and it is pitch black outside. It is a terrible situation with planes so computerized.

  9. Avid reader that hates to pile on but the analysis presented does seem rather simplistic: Vet pilots that forget the basic rules of flight.

    A couple of variables for you / others to react to:
    – It seems like the pilots did give stick down input at least once but were rewarded with a stall warning. I’ve read that under 60 knots the stall warning turns off (in some sort of alternate state). Could it be that the pilots didn’t realize they were in a stall / reacted to inconsistent feedback from the plane given the above (stick down = stall warning / stick up = no stall warning given under 60 knots). hence they kept giving stick up
    – the A330 has no AoA indicator

    1. It seems like the pilots did give stick down input at least once but were rewarded with a stall warning. I’ve read that under 60 knots the stall warning turns off (in some sort of alternate state). Could it be that the pilots didn’t realize they were in a stall / reacted to inconsistent feedback from the plane given the above (stick down = stall warning / stick up = no stall warning given under 60 knots). hence they kept giving stick up

      It appears that the when the pilots gave the stick down input, they arrested the 7,000 feet per minute climb rate they had caused and settled around 37,500 feet with only a 4% angle of attack. The next time we hear about a stall warning, it says this:

      “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.”

      So the stall warning wasn’t triggered again until nose-up inputs were again being given. And being up over 37,000 feet seems like a dangerous place to keep giving nose-up inputs considering they weren’t comfortable going above 35,000 feet before that.

  10. As in the movie “Apollo 13” (on steroids), the crew was faced with a cascading number of messages at or near the time the pitot tubes clogged, the autopilot disengaged and the air data computers (ADC) became “confused” and the airplane transferred into “Alternate Law”, likely providing mixed and/or erroneous signals to the primary flight instruments (all of which are driven by one or more of the redundant ADCs)…

    With the benefit of 20/20 hindsight, the only reliable instruments were the mechanical standby attitude indicator (peanut gyro) which has a separate 30 minute power supply, and the mechanical altimeter and airspeed indicator. I suspect that if they had received instructions to “fly the peanut gyro”, they likely would have had the information necessary to continue controlled flight.

    In the “heat of battle” when faced with multiple confusing inputs that have never been seen before, many people under such extreme stress revert to basic instinctual actions (pull up to avoid the ground) contrary to training and logic…

    IMO, today’s computerized airplanes are marvels of science and technology, but have removed the pilot from the “flying” to the “managing” role on the flight deck…

  11. I am certainly with Endre where he describes the situation in the cockpit (dark, in a cloud and inconsistent instrument readings.) Alarm lights go on, alert horns blare… Most of your time flying the A330 has been on auto-pilot and you hand-fly in “normal law” for a few hours per year, mostly at low altitude during take off and approach.
    Now you’re pushed back into flying the plane in “alternate law”, probably the first time outside of the simulator and some of the control inputs you put into the plane act differently… you can stall the plane now.

    More hand-flying gives the pilots a better feel for handling the plane (and I recommend to require some cruise-level hand-flying every month.) But that does not address the issue that pilots should have some practice in controlling the plane in alternate and direct laws…

  12. Actually it is completely false to say turbulence did not play a role in this matter. Turbulence causes disorientation. When in turbulent weather, it becomes extremely difficult to disassociate turbulence with what the plane is actually doing…..particularly in zero visibility conditions.

    When the aircraft is severely bumped around and you have no visual reference and instruments begin to fail, you may not even know or realize that your aircraft is in an usual attitude or perhaps even stalled. There are occasions where turbulence got so bad while flying a Cessna 152 that the stall warning was chiming on and off constantly due to radical airspeed fluctuations…..can you imagine flying in such conditions in the dark and with no visibility?

    So this author is 100% wrong to think turbulence didn’t play a role.

  13. And to say the pilots screwed up is ridiculous….you have no idea what those pilots were going through and cannot possibly comprehend the severity of the situation sitting in your office writing this article.

    Sounds like you’ve just got a beef with the airline/aircraft/and these pilots. RIP.

  14. The one thing we know for sure is that there will be changes to pilot training because of this, probably across all aircraft and all airlines. In the end we will all be safer on future flights because the black boxes were found and the questions about what happened were answered. While this fact is likely little solace for the relatives of the 228 people lost, I contend that their deaths are not in vain.

  15. Sorry but flying an a330 at FL350 is not the same as tooling around in a C172 at 5000 feet. You may only have a +/-5kt margin to keep the plane inside, and one really does rely upon the a/c to give an indication of the true speed. Do you really think any pilot can give you the airspeed to within 5 knots even in smooth air if blindfolded?

    In the low/slow regime such as approach and landing, your stall graphic is true. At cruise in a highly turbulent regime, not so much.

    Stall does not necessarily mean that your AoA is too high. It means that airflow has separated from the wing. Under normal flight condition, yes stall = AoA too high. Under AF447 conditions? I’m not so sure. I can stall a wing at 0 AoA in the wind tunnel by setting up a high adverse pressure gradient.

    1. I’m not trying to say that the aircraft fly the same…you are missing my point completely.

      What I’m saying is at the moment when things are occurring, it’s extremely difficult to assess what’s happening.

      A high AoA is relative to the flightpath…..and a stall occurs when separation occurs….which is always due to the angle of attack exceeding what is allowed for the design of that wing.

      1. Sorry, I was commenting on the original image of the Stall check and cross image. Yes, it’s highly confusing, which is precisely why its so easy for the a/c to move outside the flight envelope.

        Separation can be caused by more than just a high AoA, it’s just the most common reason. You can effectively stall a wing at 0 AoA by setting up an area of high pressure behind the wing, entirely possible in a weather event. If you sketch out the vectors this is effectively an increase in AoA without *actually* increasing it.

  16. I fully understand that pilot error can be a contributor to an accident. I doubt anyone is denying that….but to assume that this was a case of these pilots not knowing what they were doing every step of the way is arrogant and facetious.

    Until someone is put in that particular position where their lives are truly at risk of ending and you are faced with your mortality, it’s pointless to lay blame…particularly in such a peculiar situation as this.

    I do believe there was an element of disorientation that occurred in the cockpit. It was a perfect storm of situations that led up to this few minutes of horror. Even with the best equipped aircraft, flying in the dark, in bad weather, and with no reference is unnerving at best and can easily turn into a tragic situation if something goes wrong.

    There are few flight regimes where aircraft can become truly unrecoverable…one of these regimes is what seems to be where AF447 ended up.. a pitched up nose, rather flat stall, and zero elevator authority.

    In this particular flight regime, the options for recovery are extremely limited….1) have enough power to increase your forward velocity until the wings regain lift and rudder/elevator authority are regained or 2) be fortunate enough to have some sort of aerodynamic incident (gust of wind, a change in center of gravity, etc) change the aircraft’s pitch where again rudder/elevator authority are regained.

    From what I understand, the aircraft was banking side to side as it was coming down….partly due to the pilot corrections but I’d venture to say that in this particular flight regime the aircraft has lost lift to the point where aileron authority has been severely degraded and the aircraft is falling like a leaf oscillating back and forth from high/low wing tip angles.

    I wasn’t there (and never ever hope to be in such a situation) so I can’t say with utmost certainty. But I feel the initial disorientation of the pilots put the aircraft in this flight regime. The power required to thrust its way out simply isn’t there on any commercial airliner. The option left was sheer luck after falling into this predicament.

    The cheapness and greed of the airlines to not install even more redundant systems in order to prevent such things leads many times to such occurrences happening. While modern aircraft do have redundant systems, even one additional redundancy could have prevented autopilot kicking off making the pilots essentially what seems to be flying somewhat blind.

    1. “The cheapness and greed of the airlines to not install even more redundant systems in order to prevent such things leads many times to such occurrences happening. ”

      Seriously?? Do you really think it’s the airlines saying ‘no, it’s alright, we don’t need a safe aircraft’? Several of the issues were bad luck, which you can’t do anything about. And the airlines have essentially no input in aircraft design other than general range/capacity/comfort/efficiency concerns, so such redundant systems are designed and included by the manufacturer, who knows what they are doing.
      What would you propose? A windsock behind the tail?

      1. Yes I do…it’s in the best interests of the airlines to push the safety envelope for profit…what makes you think they operate on a higher level than any other corporation trying to milk profits?

        There are PLENTY of examples of how airlines resist change and do not want to spend money to improve safety….HELLO SOUTHWEST.

          1. Likewise for you too…oh yes…I forgot…you’re the guy who needs to get the last word all the time right

            Anyway, I’m not going to waste my time on rebutting your nonsense this time around. Unless you have something factual to say, keep to yourself please.

  17. Of course airlines have saying what to install not to install. American decided no to install Right side stick shakers in their DC-10s to reduce cost. When the engine fell off on AA-191…Investigators found that as the jet was beginning its takeoff rotation, engine number one on the left (port) wing separated and flipped over the top of the wing. As the engine separated from the aircraft, it severed hydraulic fluid lines and damaged the left wing, resulting in a retraction of the slats. As the jet attempted to climb, the left wing aerodynamically stalled while the right wing, with its slats still deployed, continued to produce lift.—The number one hydraulic system, powered by the number one engine, failed but continued to operate via motor pumps which mechanically connected it to hydraulic system three. Hydraulic system three was also damaged and began leaking fluid, but maintained pressure and operation up until impact. Hydraulic system two was undamaged. The number one electrical bus, whose generator was attached to the number one engine, failed and as such several electrical systems went offline, most notably the Captain’s instruments, his stick shaker, and the slat disagreement sensors. While a switch in the overhead panel would have allowed the Captain to restore power to his instruments, it was not used. In any event, the First Officer was flying the airplane and his instruments continued to function normally— The First Officer HAD NO STICK SHAKER INSTALLED!!!! –The First Officer’s control column was not equipped with a stick shaker; the device was offered by McDonnell Douglas as an option for the First Officer, but American Airlines chose not to have it installed on their DC-10 fleet. Stick shakers for both pilots became mandatory in response to this accident….

    1. I’m curious how relevant a 30 year old crash really is. Airliner safety has gotten much better, so isn’t it possible that airlines son have these options?

      From what I’ve seen on the sidelines airliners seem to be much less customizable right off the line.

  18. there is a lot of confusion on this thread…angle of attack is not the same as pitch angle…the A330 does not have an angle of attack indicator…AoA is a percentage of actual lift vs total lift on a wing. AF447 was at an AoA of 40 degrees BUT the angle nose up was only 16 degrees (if you carefully read the BEA report on the FDR). so this story is not simply the pilots stalled the plane…besides transport category swept-back winged jets usually are un-recoverable from deep stalls, and if the pilots saw a pitch angle of 16 nose-up, full power, unknown airspeed, and vsi showing -10.900 ft per minute, it may not be instinctive to push the nose down even more..(or there may be insuficient airflow over the elevator-making them useless).. .on older a330’s there are times when the screens cannot keep up with data changes and only show a message that says=”invalid data”…signed d.p.-3200TT a330, 8700TT a320

    1. if the pilots saw a pitch angle of 16 nose-up, full power, unknown airspeed, and vsi showing -10.900 ft per minute, it may not be instinctive to push the nose down even more

      But the pilots weren’t presented this situation. They got themselves into it. When the autopilot shut off, that’s when the nose-up commands started and they ended up in this situation.

    1. I certainly hope so. This is not a complete report – it’s an attempt to boil down the main points into something for the public. When the full 200+ page report comes out, it better be more complicated but I expect the basics to stay the same.

  19. There is a published procedure in the Airbus manual…3 degrees nose up and 83% N1 engine speed maintains altitude and airspeed until the pitot tube heats enough to dislodge the ice…this has happened on other Thales tube equipped 330s….this is not to blame the AF447 pilots who no doubt had many other issues including reports of severe turbulence in their area.

  20. Pardon my ignorance as I try to understand this, but it seems like the pilots essentially…what, overcompensated?

    1. Great question. I don’t think I’d say they overcompensated but possibly mis-compensated (even though that’s probably not a word). The movements that the pilots made were not consistent with what should have been done in that situation.

  21. WHAT the heck are they stalking about: not enough training? I am only airplane single engine land, but during training about 60% of it was practicing how to recognize and recover from a stall. When air speed falls, WHO in their right mind would RAISE the nose of the airplane? Even if airspeed in increasing, that would not make sense unless something indicated the plane was in a dive. WEIRD.

  22. All the indications are that a severe up draft pushed the plane up to 38,000 and into a high pitched up attitude and a stall which the pilots did not have enough flight control to deal with.
    Every operation manual I’ve dealt with for 40 years states “Avoid thunderstorms”.
    You can guarantee the pilots were unable to deal with this, not unwilling.

    1. I have seen no such indication and that’s inconsistent with the inputs given by the pilots. If a severe updraft somehow pushed the airplane up 3,000 feet, then the pilots knee-jerk reaction would have been to make nose-down inputs. Instead, they made the nose-up inputs themselves.

      1. @CF:

        You think you know too much….but you don’t. You really don’t what happened and you’re basing your assumptions on the rather thin amount of released data and media coverage that’s been given thus far.

        It’s extremely presumptuous to assume they made “knee-jerk” reactions and it’s stuff like that which spreads false rumors and suppositions about truly occurred.

      2. How do you determine that the pilots made “knee-jerk reactions?” Or that they continued to raise the nose?
        Are you implying that they suddenly decided to zoom climb up to 38,000 feet when they knew that they could barely maintain 35,000 feet? Why climb at 7000 fpm?
        If they planned the climb why no reference it on the CVR? Did they clear it with ATC? Something caused the plane to climb, but it wasn’t the pilots.

        1. I’ll say it again. It’s quite easy to know that the pilots caused the climb because the data recorder measures inputs, not just aircraft movements. As soon as the autopilot disconnected, the Pilot Flying made nose-up inputs and those were consistent through most of the rest of the next few minutes.

          I never said anything about “planning” to climb or anything along those lines at all. I’m simply reviewing the data and we know that the pilots consistently tried to point the nose up, at least until the captain came back.

          If you’re going to keep citing severe updrafts as your culprit, please include a link showing where you’re getting this because I have seen nothing anywhere to indicate that.

  23. “But now, the French probably are wishing those black boxes remained on the floor of the ocean because its national airline is about to face some tough questions”.

    Or probably not. Either Air France or French co-owned Airbus – main factory in Toulouse – and probably both, will have to face tough questions anyway. “The French”, who lost many coutrymen in the crash, would rather try to understand what went wrong.

    1. Excellent link….some people should actually read that article and stop making such ridiculous “factual” assumptions about what they think happened.

    2. I saw Patrick’s piece; it’s a good one. But really it doesn’t say much different than what I’ve said. I put more emphasis on the incorrect control movement from the pilots but as I said at the end, there was a lot going on at the time. There isn’t a simple explanation because it’s always a bunch of factors that lead to an accident.

      Here is a good post from a Delta A330 pilot that I think gives a lot of great information. She puts plenty of blame on the pilots but also talks about how the training systems really need to change. Some of you will be particularly interested in seeing the part where she suggests using GPS groundspeed to help in that situation.

      And for those who care, “Noz” is the same person as “Armondm” just using a different name. I have no plans on responding to his angry rants myself, but I thought others might like to know.

      1. Angry rants?

        I’m not angry at all…I just find it disgraceful on your part to squarely put the blame on the pilots based on media BS and incomplete knowledge of what went on during that tragic period of time.

        Now you’re backpedaling with your “There isn’t a simple explanation…” comment…nice.

      2. Really? Patrick Smith seems to say there’s no easy answer as to what happened, while your post is (in my interpretation) trying to provide an easy answer. To quote his post:

        “It is easy to look at the early report, with its voice recorder transcript and summary of control inputs, and come away with a seemingly clear picture of the crew making a bad situation worse, to the point of catastrophe. They pulled up when they should have pushed down. But was it really that simple?”

        I am not a pilot and have no idea if you are right or wrong, so it’s pointless for me to speculate what happened. But I hope you don’t mind if I put a bit more trust into the blog posts of an actual airline pilot than that of a well-respected (by me) airline geek :)

        As for that Delta A330 pilot… geez, she really doesn’t seem to like Airbus aircraft based on some of her comments. Why does she fly them?

        1. Mine is certainly not providing an easy answer. As I said in my first paragraph:

          No airplane accident happens because of just one problem, and this is no exception, but so far pilot error is really sticking out as the single largest contributor here.

          And I absolutely stand by that. (And I am not backpedaling in any way, as had been suggested by another commenter.) But that doesn’t mean it was the only contributor. Patrick talks a lot about stalls.

          How the pilots of Flight 447 found themselves on the verge of a stall in the first place, and why they reacted the way they did — apparently pulling the nose up rather than pitching it forward, creating a series of worsening oscillations that ultimately resulted in a full, non-recoverable stall — is not yet clear. It may never be understood fully.

          I don’t ever suggest that I know why the pilots did what they did, just that they did make inputs that ultimately made things worse and brought the airplane down.

          In the end, Patrick says this:

          Was it failure of the data sensors? Was it violent turbulence? Was it pilot error? I suspect it was a combination of those things — an unlikely but deadly cascade.

          I still haven’t seen any indication that they encountered severe turbulence, so I will admit to disagreeing with him there. But it undoubtedly was a chain of events. Would the plane have crashed had the pitot tubes not frozen and the aircraft continued operating as normal? Probably not. Did the alarms distract the pilots and make them act incorrectly? Possibly. Is training inadequate to help pilots deal with situations like these? Could be. It’s never just one thing.

          But in this case, the actions of the pilots stand out as being the largest contributor for me. That, of course, doesn’t mean anyone has to agree with me on it. Even when the final report goes out, we’ll never know the reasons behind everything and there will continue to be disagreement.

  24. Those of us who fly live by one cardinal rule: “Fly the airplane first, worry about anything else second.”

    No doubt pilot error will be found in this case, even if it was unavoidable because of the loss of proper inputs into the computer.

    On a different note, it makes you wonder if pilotless aircraft will really catch on – after all, can we really trust computers who have been programmed by human beings who may not think of a problem that might arise?

    1. That’s a fascinating question, I think. It’s an even more interesting question around whether we would even go to single pilot scenarios for commercial aircraft. There are plenty of reasons against it, but then again, nobody thought we’d ever stop having a third person in the cockpit.

      Ultimately, automation can’t solve everything and accidents will happen, but if it can be shown to be incredibly safe, then I imagine one day we might see it. Certainly the airlines would be happy to consider it because it would reduce costs, but only if it was safe enough. The biggest hurdle would (and should) be the government being willing to sign off on it.

      I don’t see it anytime soon, but it wouldn’t surprise me to see it in my lifetime.

    2. It’s a scary thought but consider adaptive algorithms that are now being used in advanced robotics and such have come a long long way.

      Of course…any error made on the human side will translate to what the computer does…unless the adaptive skills of the computer can “recognize” something wrong and correct it.

  25. No evidence of turbulence??? They flew dead on into a thunderstorm. Why did everything start to fail? What caused the icing and the failures of the most necessary instruments? What caused the plane to zoom up? Why did every other plane in the area deviate around the CBs? If there was no flight into the heart of the storms, then none of this would have happened.
    Thunderstorms can be very dangerous!
    Every operations manual warns: Avoid Avoid Avoid!

  26. Well, if a human error is involved in the crash (or at least partially involved) I’d really like to know if the relatives of the victims of the crash filed a lawsuit against the company. It’s the right thing to do in these cases, in my opinion. I know the company is a respectable one, but the relatives of the victims should get some comfort and air companies usually get by just fine after such accidents.

Leave a Reply to cstclair Cancel reply

Your email address will not be published. Required fields are marked *

This site uses Akismet to reduce spam. Learn how your comment data is processed.

Cranky Flier