TITANIC of the Skies! - The Untold Story of Air France 447

(rain patter) (plane engines roar) (plane crash) (somber music) - It took just over four minutes for this plane to lose control and fall almost 38,000 feet. That's about the same time it takes to toast a piece of bread. What were the crucial details that caused this accident? Stay tuned. The crew that was to operate Air France Flight 447 arrived in Rio de Janeiro from Paris on May 28, 2009. They had planned a three-day stopover in Rio before the return flight to Paris. We know very little about how they spent their time during this stopover but what is known is that the first officer who was later to be the pilot of this flight had brought his wife along for the trip.

Given this, it is unlikely that he spent at least the entire day before the flight resting and there are indications that this could also apply to the other pilots. Flight 447 was scheduled to take off at 8:00 p.m., Rio time, or 11:00 p.m., Paris time. And that meant the crew would have to operate all night back home with a scheduled landing in Paris around noon on June 1. This would mean that pilots would have to operate throughout the low circadian rhythm window, a time of night during which it is especially difficult for the human body and mind to stay alert.

Due to this and the fact that the flight was scheduled to last 12 hours and 45 minutes, the flight crew had been augmented with an additional pilot to allow for scheduled rest during the return flight. This meant that there were three pilots who were going to fly together: a captain and two first officers. The captain was 58 years old and had a total flight experience of 11,000 hours, 1,700 of which he flew as pilot in command of the Airbus A330. The first officer was 32 years old. He was the least experienced of the crew with 3,000 hours total and 800 on type. He had type-rated him on the Airbus A330 and 340 in 2008, the year before the flight.

The third pilot that I will refer to in this video as a relief pilot was a 37-year-old first officer. He had a total time of 6,500 hours, but in reality he was the most experienced in this type with 4,500 flight hours. During the three months before the flight he had flown very little because he also worked as a management pilot at Air France. Prior to operating the flight, the pilots met and reviewed planning documentation, including weather, flight plans, and NOTAMs. The weather was good for the departure and in Paris, but the weather on the road seemed a little more complicated.

Since the flight would cross the equator, it would also cross an area over the Atlantic known as the Intertropical Convergence Zone, or ITCZ. The eastern trade winds from the northern and southern hemisphere will converge in this area and can push moist air from the sea surface upward. This can cause powerful storms that can reach up to 60,000 feet, which is much higher than most airplanes can fly. This also means that planes that have to fly through this area could be forced to fly between and sometimes even through parts of these storm clouds. Storms that form over water in the ITCZ ​​can sometimes have a lower water content than storms that form over land, making them more difficult to see on aircraft weather radar, so pilots must Be very careful when navigating around them, as they could still be quite severe.

The pilots probably discussed the risk of encountering these types of storms and carried some extra fuel to give them the option to navigate around these storms if necessary. The pilots also informed their nine cabin crew about the possibility of turbulence and the effect it could have on their service. Before we leave the discussion about the Intertropical Convergence Zone, we need to talk a little about the types of precipitation that can be found within these types of storms. Most people are familiar with the heavy rain and sometimes hail that typically comes out of these clouds, but as you look higher up through the clouds, things can get a little more complicated.

Due to strong currents within the cloud, water droplets can be pushed upward into very cold air and become very cold. This means that they are still in water form, but as soon as they hit the surface of some type, they will freeze instantly and create clear ice. Sometimes these super-cooled droplets can collide with snowflakes and when that happens, a kind of soft ice crystals can form, which are not as hard as hail but large enough to be heard when the plane hits. if you fly through it. This type of precipitation does not create the type of heavy fuselage icing that supercooled rain creates, but it has significant volume and can quickly clog and overwhelm the aircraft's sensors and probes, especially the Pitot probes, which I will explain later. .

During 2008 and 2009, more than nine different cases of this type of obstruction were reported on Air France flights. And these reports, along with how to recognize and address problems, were published in safety bulletins distributed to all Air France pilots in the year before the accident flight. The plane the pilots were going to operate was a reasonably new Airbus A330-203. It was delivered to Air France in 2005 and was in perfect working order on the afternoon of its departure. Before reaching the plane, the pilots ordered 70.4 tons of fuel to be loaded and, together with their cabin crew, walked to the plane and began preparing it for departure.

Now it will be impossible to explain what happened on this flight without also explaining some details about this Airbus but also a little about the flight in general. Pilots need a way to accurately measure the amount of air flowing over the wings because that's what really determines the aircraft's performance. To do this, they use a type of probe known as a Pitot tube or Pitot probe. These probes are usually located at the front of the plane, under the cockpit and look a bit like gun barrels. And they have a hole in the front where the air enters and then the total pressure inside the tube is measured.

They are heated electrically and the heating is automatic on the Airbus A330. But to accurately measure air velocity, static pressure must also be measured so that it can be deduced from the total pressure of the Pitot probes. This static pressure is measured from a different device called a static port and then that static pressure is used to both calculate the airspeed and, crucially, also the altitude of the aircraft. These different pressures are then sent to the aircraft's computers, called Air Data Modules or ADMs. The ADMs will calculate the correct actual airspeed, but another thing that will be hugely important in this

story

is that the static port static pressure needs to be corrected depending on how fast the plane is flying.

This is because air flows over the surfaces of the aircraft, surrounding the static port and will therefore create localized pressure differences depending on speed. These corrections are made automatically by ADM computers and because airspeed and altitude are critical values, there are three different and independent sets of probes and computers installed on the aircraft. Now, because Air France and other operators had reported problems with ice crystals clogging the Pitot probes, Airbus had begun investigating the problem. A newer type of Pitot probe was found to be more effective in preventing these problems and a maintenance bulletin was issued suggesting an upgrade to these newer probes.

Air France had just begun modernizing its first Airbus A330 about a month before the departure of Flight 447. And the crashed plane's probes were scheduled to be changed upon arrival in Paris after the flight. But why was changing these Pitot probes just a suggested action? Why wasn't it mandatory? Well, that's because the temporary loss of airspeed due to this problem was very rare, it only lasted a maximum of a couple of minutes and there was a defined procedure that the pilots had to follow in case it happened. At Air France, this procedure was known as IAS douteuse, but in this

story

I will refer to it as unreliable airspeed.

Because this problem had been reported several times, it was included in the recurring training scenario for all Air France crews during 2008 and 2009. The training included unreliable airspeed exercises, but only at low altitude. This is because it was considered more critical for safety due to the closeness to terrain, but the aircraft's performance was also much better than at high altitude. Unreliable airspeed can be very difficult to diagnose because the fault will look different depending on the cause and severity of the fault. During the exercises that the Air France crews had practiced, the autopilot did not disconnect and no warnings were heard in the cockpit when the failure occurred.

Now, the unreliable airspeed procedure included the use of rote items, i.e. safety-critical items that had to be performed immediately from the pilots' memory. But using them was optional depending on the situation and had been interpreted as necessary only if the plane was close to the ground. Also important to this story is that none of the pilots on Flight 447 had received recent training on how to handle a stall approach and recovery, especially at high altitude. The last stall training they had received was during their type rating on the Airbus A320, which they had all completed years before.

And that initial training that they had done was done at low altitude in which great emphasis was placed on using thrust to recover the airplane and get it out of the stall, accomplishing it with minimal loss of altitude. Toning down was a secondary action to take. Now, this idea that the engines will have enough power to get an airplane out of an extremely high angle of attack is also going to be very important. Once all 216 passengers finished boarding and the crew was ready, the pilots called back and began pulling away from their gate at 22:09 UTC, just nine minutes late.

The co-pilot was the pilot at the controls and all three pilots were present in the cockpit while filming. At 10:29 p.m. the aircraft began taxiing on the Rio de Janeiro runway and took off completely normally. They headed off, following the clear exit route that took them northeast, following the coast of Brazil. At some point, after exceeding 20,000 feet of climb, the relief pilot left the cockpit to begin his scheduled rest period that would last about three hours. He returned to the crew rest compartment which consisted of two bunks just behind the cabin and we don't know exactly when he left the cabin because the voice recordings from the cabin voice recorder don't start until shortly after midnight, UTC. time.

In any case, the two remaining pilots were cleared to climb to their initial cruising altitude of flight level 350 or 35,000 feet and once established in cruise, they spoke with the Brasilia FIR controllers and then changed to Recife Control. . And it was Recife Control who were going to be the last to have radar contact with the flight because in that control area, they were largely still flying over land, but once they moved to the next FIR, Atlantic, they would move away from the Radar. coverage and instead follow ocean traffic separation procedures. Today, flights over oceanic areas require special training procedures and aeronautical equipment.

Since the curvature of the Earth makes VHF radio communication impossible, airplanes are equipped with something called HF radios. These radios use layers of the ionosphere to bounce signals and can therefore reach much greater distances. When the aircraft passed an RNAV point called INTOL, the pilots checked in with Atlantic Control on one of the two HF frequencies they had been assigned. The pilots then attempted to log into a new system being tested at the time in the area called ADS-C. This system would use automatic reports sent by the aircraft itself via satellite to update the position of the aircraft to ATC, thus showing where it was even if they did not have radar coverage.

Another interesting thing this new system could do was immediately send a report if a plane deviated from its assigned heading or altitude. But unfortunately, due to a formatting error in the flight plan that had been submitted, the pilots were unable to log into this new system and therefore, as soon as the aircraft left conventional radar coverage, it would no longer it would be possible to follow it precisely. , something that would have serious consequences. The Airbus A330 is a fly-by-wire aircraft and that means that the inputs that pilots make on their side sticks and rudders will be interpreted.electronically by a computer and then sent to the hydraulic flight control actuators for execution.

This type of controls has many benefits, for example it makes the aircraft substantially lighter but mainly it allows you to monitor certain safety parameters and ensure that they are not exceeded. Parameters such as excessive tilt and tilt angles are monitored, in addition to safeguarding maximum and minimum speeds and a host of other parameters as well. Now detractors of the system say that this allows the plane to have the final say over the pilots, but that is not entirely true. The system only blocks maneuvers that are really extreme and ultimately dangerous, but the fly-by-wire system requires the pilots operating it to really understand how they work and when those protections actually work and don't.

Obviously, this is true for all airplanes, but it's especially true here. And why is that? Well, for these protections to work correctly, the computers who monitor them must be absolutely sure that they are using the correct parameters to begin with. If that's not the case, computers will go back and remove those protections, simply because they're not really sure what's going on. These computers receive their inputs from many different sources, such as pitot probes, static ports, inertial reference units, angles of attack, etc. They combine all of this data into three Air Data Reference Units that together form the Air Data Inertial Reference System.

As long as all three or at least two of these ADRs match each other, the plane's control computers will be happy and it can continue operating in what is called Normal Law. Normal law means that all protections are available and that it is basically impossible for the aircraft to stall or get into a disturbance situation. But if two or more ADRs start sending strange information, a couple of things will quickly happen. First, these inconsistencies could affect the automatic flight system, such as the autopilot that controls the aircraft, the autothrottle that governs the engines, and the flight directors that show pilots how to fly.

That's pretty logical, if you think about it. The aircraft will not attempt to navigate or control the aircraft if it is not sure what is happening. That will be up to the pilots to resolve. And following that same logic, the aircraft's control computers will degrade from Normal Law to Alternative Law 1 or Alternative Law 2, depending on the severity of the problems. The difference between Normal Law and Alternative Law 2 that will soon become relevant in this flight is that the protections that the aircraft normally has in terms of maximum angle of attack or stall protection will no longer be available.

This will be demonstrated by the removal of warning indicators such as the barber pole on the main flight screen, as well as the yellow crosses where limitations would normally be displayed. The other difference is that it changes the roll control of the airplane. In Normal Law and Alternate Law 1, the roll inputs on the side stick will command a specific roll speed of the aircraft. If the pilot enters a specific roll rate to be maintained, the aircraft will give it and any gust disturbances will be compensated. Basically it will be very stable and easy to handle. But in Alternative Law 2, the side stick will give direct commands to the ailerons and spoilers instead of commanding a specific roll rate.

This means there are no bank protections or stability control. This will also make the aircraft more sensitive to roll, especially at higher altitudes where there is less aerodynamic damping due to thinner air. Another important difference between a conventional aircraft and an Airbus fly-by-wire aircraft is the pitch compensation system. If you fly manually in a conventional airplane, the yoke controls the flight controls directly and pilots must make the adjustment deliberately. On an Airbus, input from the side stick will request the control computers for a specific roll speed horizontally and a pitch or g-load vertically. When the pilot sets a specific pitch, the elevators will initiate the pitch and the massive horizontal stabilizer will automatically move to continue maintaining that pitch without any input from the pilot.

This also means that there isn't much tactile feedback from the stick if the airplane goes into an odd trim position due to low speed, for example, which would be the case on another airplane like the 737, for example, which I'm flying. . Under normal law, that's not a problem because it also prevents the plane from approaching a high enough angle of attack to stop it, but that's not the case under alternative law. Now, it is really important for any pilot to know these differences, but what is even more important is that they must understand them and be able to retrieve this knowledge when things start to go wrong.

Now before we get into the actual accident sequence in this video, I just want to share this quick message from my sponsor who allows me to create these videos. This is really cool! Are you looking for ways to save a little money on your next vacation or online shopping? Well, did you know that you can get cheaper plane tickets, hotel reservations, or even car services by connecting to a VPN in a different country? By shopping online with current sponsor NordVPN, you can find the best deals in the world no matter where you buy from because it turns out that prices actually change, for example, on booking.com when you connect from a different country.

I have no idea why, but thanks to NordVPN, you can save a lot of money by trying it. Nord has over 5,500 ultra-fast servers in 59 different countries, so there's plenty to choose from. So whether you want to watch your favorite content from other countries, protect your personal information, or save money with a VPN through regional discounts, NordVPN has you covered. Start surfing the Internet freely and safely today and right now, if you visit NordVPN.com/pilot, you'll get an amazing exclusive offer. And you really have nothing to lose because Nord always offers you a 30-day full money back guarantee.

Thanks NordVPN. Now let's go back to the video. As Flight 447 continued on its route, they received a message from their company, informing them of the storm activity reported ahead. The first officer changed the scale on his navigation screen and began talking about some accumulations he could see ahead of them. The pilots then began discussing ways to possibly avoid the turbulence these clouds would bring. And the co-pilot indicated that he was quite concerned about the storm clouds that he now saw approaching. At 01:46 the co-pilot lowered the intensity of the cabin lights to be able to see the outside better and confirmed that the aircraft was already entering a layer of clouds.

It was pitch black outside with the moon situated behind them and slight turbulence now began to rock the plane. The crew discussed whether it would be possible to climb higher, as it appeared they were simply skimming the cloud tops at flight level 350. But when the first officer checked the maximum altitude they could climb, it was listed as flight level 370. and they both decided that it would not be a good idea to sit at maximum altitude if the turbulence got worse. Now, this is probably a good opportunity to discuss something else that will be important very soon.

As a plane climbs higher, the air outside will become increasingly thinner. And because of that, depending on the weight of the aircraft, as well as the outside temperature and some other factors, the aircraft will not be able to continue climbing as the engines simply will not be able to produce enough excess thrust to do so. This also means that the engines don't have much excess thrust to deal with anything abnormal up there, and so we pilots have to be very careful and quite conservative with how the plane is operated. The only way an airplane can climb above its maximum indicated altitude would be to exchange kinetic energy, its speed, for potential energy, altitude.

But doing so would dramatically increase the angle of attack. At higher altitudes, the critical angle of attack, that is, the angle at which the aircraft will stall if exceeded, is also lower due to the aerodynamic effects of flying closer to the speed of sound. In short, the plane will stop sooner and have less thrust available at high altitudes. It will also be more sensitive and harder to fly by hand because there is less air providing aerodynamic cushioning. Once the plane entered the clouds, the captain began to comment on the appearance of the San Telmo fire. This is a sure sign that electrical charge is building up near a plane, one more indication that they were approaching storm clouds.

At that moment, the captain pressed a button on the overhead panel, which activated the call bell in the crew berth behind the cabin. This was the signal to the relief pilot that it was time to return to the cockpit and replace the captain so that he could return to the bunk and get some rest. The relief pilot knocked on the cockpit wall in response to demonstrate that he had heard the bell. Now, Air France's standard operating procedures at the time stipulated that the first officer in the right seat who was the pilot in command would assume the role of pilot in command when the captain went to rest, unless something had been agreed upon. further.

And in this case, this is exactly what happened and when the relief pilot arrived in the cockpit at 01:59, the captain stood up from his chair and listened to the briefing that the co-pilot gave to the relief pilot. This briefing included the expected turbulence, that they were currently flying in a cloud and could not go any higher at this time. The captain left no specific instructions as to how the two first officers should confront the coming storms. Instead, he left the cabin to start resting. Now we can only speculate as to why he didn't give clearer instructions.

As I mentioned before, all three pilots were used to this type of weather and had flown this route many times. Therefore, it is likely that he found the first two officers more than capable of coping. When the captain left, Air France procedures meant that the much senior relief pilot now assumed the role of follow-on pilot in the left seat and the first officer continued as flight pilot and, effectively, lead pilot. command. The two pilots began to discuss the weather situation ahead. And the first officer mentioned the Intertropical Convergence Zone. But at this stage, there was no discussion about trying to avoid the cells by changing course.

Instead, the first officer asked the relief pilot if he had been able to get any sleep and the relief pilot said that he had simply fallen asleep a little. The relief pilot then asked the co-pilot if he was feeling okay. This discussion could be interpreted as the two pilots possibly feeling a little fatigued at this point. The captain had also made similar comments earlier. The relief pilot also left his pilot's chair in the aft position and did not move it forward to the piloting position when he sat down. That probably would have given him a more relaxed position where he could put his feet up, but it wouldn't be ideal if he suddenly needed to fly the plane, which he would soon need to do.

At 02:06:05, the co-pilot contacted the cabin crew to warn them that they would likely encounter slight turbulence soon and to tell them to be careful. About four minutes had already passed since the captain left the cabin. And in another four minutes the emergency sequence would begin. The relief pilot began to look more closely at his navigation screen, crouched down, and changed the weather radar gain from him to Max. This made the weather radar more sensitive and highlighted the weather ahead of them. He then asked the co-pilot if he perhaps wanted to go a little to the left, indicating some discomfort with what he was seeing in front of them.

The first officer asked what he meant and then pointed to the turnaround on the screen in front of them and asked again if they could turn left. The first officer agreed and turned course about 12 degrees to the left. This was a fairly small correction given the amount of weather in front of them, but it meant that they now abandoned their flight plan route and did not attempt to radio this change to air traffic control. At 02:08:17, the cockpit voice recorder detected a change in background noise in the cockpit. This noise sounded like rain hitting the cabin. At the same time, the co-pilot noticed that the temperature was rising and asked the relief pilot if he had done something with the air conditioning.

He then asked, "What is that smell?" The relief pilot recognized the smell of ozone and calmed his colleague by explaining what ozone was. These comments, as well as thePrecipitation that was heard outside the plane were signs of the type of meteorological phenomenon that the plane was flying towards. The change in temperature and humidity felt in the cabin was probably another sign that the engines suddenly absorbed a large amount of water or ice, partially overwhelming the air conditioning system. 20 seconds later, the background noise intensified and transformed into the sound of ice crystals now hitting the cabin. The turbulence intensified and the first office reduced the commanded speed from Mach 0.82 to 0.8, which was the penetration speed of the turbulence.

This caused the autothrottle to reduce thrust to 84% N1. Up to that point, the plane had functioned perfectly and the two pilots were relaxed and simply monitoring that the autopilot was doing its job. We don't know for sure what caused it, but it's very likely that the ice crystals that were now being thrown at the plane outside began to overwhelm the heating elements of the pitot probes that were providing crucial information to the three air data modules. At 02:10:05, the autopilot and autothrottle were suddenly disconnected with the associated cavalry charge warning in the cockpit. At that same moment, a strong gust of wind caused the aircraft to begin to bank sharply to the right.

The first officer yelled, "I have controls!" and reached for his side lever. Now, this sudden appearance of warnings and aircraft behavior would probably have caused both pilots a great start and confusion would have started almost immediately. But at this point the plane was still flying reasonably level and only this right bank was developing. As the airplane continued to bank to the right, the airspeed indicated on the captain's primary flight display, as well as the holding indicator, suddenly dropped from 275 knots to 60 knots, a clearly erroneous indication. We don't know for sure what the first officer's instruments showed, as they were not recorded on the flight data recorder, but they probably also fell in a similar amount.

Remember what I told you about how the altimeter system compensated pressure for higher Mach numbers? Well, because the speed was now indicated to be much lower, the altimeter system recalculated his altitude and suddenly showed a 300 foot drop as well as a negative vertical speed. The first officer who had suddenly been forced to fly the plane by hand now reacted instinctively by turning the left aileron and pitching upward to correct his altitude. This was the beginning of the sequence of events that would ultimately doom this flight. Since all three ADRs were now showing unreliable data, the aircraft reverted to Alternate Law, which meant angle of attack protections were lost, as well as overspeed and roll protections.

Now, as I mentioned before, the roll control also now became twice as sensitive, meaning the first officer had to really concentrate to get the plane to stop pitching and return to level flight. He began to overcorrect and the movements he made with the side stick began to cause more oscillations from right to left and then back again. It is quite possible that most of the first officer's attention was focused on this part of the steering at this point, but remember that he had also begun to pitch. That tone was now increasing, leading to a very high rate of ascent.

Almost instantly, as the autopilot disengaged and all the different warnings began, a brief stall warning was also heard in the cockpit. When the warning sounded, the relief pilot shouted, "What's that?" Probably due to the quick nature of the warning, but could also indicate confusion over its meaning. The crew did not disconnect the flight directors as instructed by the procedure for unreliable airspeed, but the command bars disappeared anyway due to a lack of reliable data. 11 seconds into the sequence, the airplane's pitch had reached 11 degrees nose up and the first officer was making maximum input with the left and right side sticks, still overcorrecting.

He now he too shouted: "We haven't... We haven't had a good display!" Probably referring to the lack of speed information. The relief pilot who was supposed to handle the ECAM warnings and the abnormal checklist, shouted: "We have lost speeds! And: "Alternative law protection law." He then continued reading the ECAM messages, but in In a difficult to understand and very hasty way, it would have been absolutely crucial here to clearly point out that the aircraft was in Alternate Law, that is, that it would be handled differently and that certain protections were lost. He mentioned that the self-propellant was lost. the first officer responded by asking, "Engine lever?", showing some confusion about what the relief pilot was saying.

Since the autothrottle had been disengaged when the failure occurred, thrust had entered a mode called engine thrust. locked, which basically meant that it maintained its previously selected value of 84%, the ECAM ordered the crew to move the thrust lever manually to resume thrust control, but instead the first officer pressed the disconnect button on the automatic accelerator and left the thrust lever where it was, which was at the climb stop. And because of that, the thrust now began to increase until it became climbing thrust. Everything I've said so far happened in the first 18 seconds of this emergency. 10 different ECAM messages appeared, each of which caused a ring in the cockpit.

In addition to this, the master warning light and master caution light illuminated in front of the pilots, as well as a constant C chord chime, indicating that the aircraft had left its authorized altitude. This would have been a very confusing and stressful environment for the pilots. And the fact that the unreliable airspeed was not indicated on the ECAM, this was something the pilots needed to figure out for themselves, that might have been the reason they didn't start running that procedure at that time. At 02:10:26, the aircraft's pitch attitude had reached 12 degrees nose up. The plane was now climbing at a hair-raising speed of 6,900 feet per minute, more than seven times what its normal speed would be after passing 36,000 feet.

Maintaining such a high speed meant that the plane was now rapidly trading speed for altitude. Suddenly, the flight director bars reappeared in front of the first officer, but instead of being activated in altitude hold mode as they were before, they are now activated in vertical speed mode since they were activated in the middle of a climb. . This meant that flight directors were beginning to show pilots the pitch needed to maintain their current 6,000 feet per minute climb. That definitely wouldn't have been helpful in this situation where the first officer was probably becoming increasingly overworked by the noise and workload.

And it was precisely because of this that the unreliable elements of airspeed memory included turning off the flight directors. The relief pilot now looked away from the ECAM screen, which he had stopped reading anyway, and managed to turn on the Wing Anti-Ice and it was then that he noticed the high tone that the plane maintained. He shouted: "Watch your speed!" Possibly referring to vertical speed since the speed indication had not yet returned. The first officer responded, "Uh, okay, I'm going back down," and he started leaning forward, but not far enough. This only reduced the rate of climb, but the plane continued to climb.

The relief pilot continued yelling that he needed to stabilize himself, "Get back down!" and "According to all three of you, you're going up, so go back down," to which the first officer responded, "Okay. The pitch has now been momentarily reduced to about 10 degrees, which is still going on." They gave a rate of climb of approximately 4,000 feet per minute. They now climbed to their maximum calculated altitude of 37,000 feet and at time 02:10;36, the flight directors disappeared again but the airspeed returned to the left side, indicating the correct speed of. 223 knots, that is, up to that point the plane had lost about 50 knots of its precious speed and the speed continued to decrease.

The flight directors reappeared for about a second, disappeared, and then returned, ordering a climb. about 1,400 feet per minute, which was what the plane was doing at that time. At 02:10:47, the first officer inexplicably reduced thrust to about 85% N1. And this is probably a sign of his increasing loss of situational awareness and disorientation. The relief pilot pressed the call button to try to get the captain back into the cockpit. He pressed it several times and asked, "Where is he?" He also reached out and changed the air data selector and the attitude and heading selector to the first officer at 3.

These are switches designed to change the data input for the displays and were not items covered in any checklist. He probably changed them in an attempt to try to restore the instrument that, in his opinion, was missing from the first mate. Three seconds later, the cabin stall warning activated and sounded continuously for the next 45 seconds. The fact that this warning was now sounding continuously was not verbalized or discussed by any of the pilots. Instead, the first officer reacted by pulling back on the controls, increasing the pitch attitude to 16 degrees nose-up, similar to what he would have on takeoff.

This pitch-up command now also meant that the stabilizer at the rear of the aircraft began to move to the maximum nose-up position, making potential recovery even more difficult. The plane no longer had enough power to maintain the requested climb rate. Instead, it slowly began to level off, swaying from side to side, which the first officer desperately tried to counteract with the side stick. They were now rapidly approaching a fully developed position. Now, a stall occurs when the plane's wing has moved beyond its critical angle of attack, meaning that it effectively stops being able to generate lift and the plane will instead begin to fall.

A loss can occur in any attitude. The only thing that determines when it happens is the angle of attack. And the angle of attack is the angle between the chord line of the wing and the approaching airflow. Both pilots had received initial training on how to approach a stall approach and recovery during their type ratings years ago, but would never have experienced the sensation of a fully developed stall in a real aircraft. The fact that the Airbus only has an audible warning, not a visual stall indication in front of the pilots or a stick shaker, could also have helped explain why there was no reaction to the warning, because one of the first things disappears when A human being is under stress is his hearing.

At 02:10:57, the aircraft reached its stall angle of attack and violent shaking began to shake the aircraft. This type of shaking has a much higher frequency than turbulence and feels very different. So the first officer now added all the TOGA thrust, but at this altitude, above the plane's maximum performance altitude, the engines were not able to produce enough thrust to even maintain speed, much less accelerate it. The relief pilot shouted: "Above all, try to touch the side controls as little as possible!" Showing an insight into the problems of flying in Alternate Law, but this fell on deaf ears as the first officer continued to make maximum left and right inputs.

The relief pilot also asked: "Is he coming or not?", referring to the captain. The first officer shouted, "I'm in TOGA," referring to the fact that he had added thrust that he might have thought would be enough to resolve the situation. The plane still climbed a few hundred more feet until it reached its maximum recorded altitude of 37,924 feet and then began to fall toward the sea. Now you may wonder why the relief pilot didn't take the controls? He was asking for corrections that didn't happen. Well, we'll never know the answer to that, but it probably lies in the fact that it all happened pretty quickly.

It only took about a minute to get to this point. Additionally, the fact that he could not see or feel what his colleague was doing on the side sticks, as they are not connected to each other on the Airbus, made it difficult to monitor what his colleague was actually doing. At 02:11:07, the heating elements removed the last bit of ice inside the Pitot tubes and both the left and standby airspeed indicators began operating again. The indicated speed was 183 knots, which means thatthe aircraft, the four control centers that were planned to be in contact with the flight, soon began contacting each other to check if anyone had heard anything from the aircraft.

It soon became clear that there had been no sign of the flight since 01:35 and that the crew had not checked in at the various position checkpoints they should have done. Since the problem with HF radio contact was not uncommon in this part of the Atlantic Ocean, no one noticed for the first few hours. Several calls continued to be made between the different ATC centers to verify if anything had been heard. And other planes flying the same route were asked to try to contact the flight, but without success. The same thing happened when trying to contact them through SatCom, which is a type of satellite phone, as well as EICAS messages sent from Air France.

Three hours after the accident, the flight was declared missing for the first time and three hours later the first emergency message was sent from Madrid in Spain and also from the ATC control centers in Senegal. This caused the first search and rescue planes to be sent towards the plane's last known location. On June 2, 2009, two days after the accident, the first floating debris was sighted on the ocean surface and three days later, the first remains of the victims were recovered from the sea. This was the start of a two-year search and rescue effort that included mapping all the underwater terrain in the area where the plane was thought to have gone down.

The search was divided into several campaigns with the last one beginning in March 2011. And on April 2 of that year, the remains were finally located in a relatively flat area on the sea floor at a depth of approximately 12,800 feet and six and a half a mile away from its last position transmitted by EICAS. This allowed thousands of pieces to be recovered, including both the cockpit voice recorder and the flight data recorder, which, miraculously, were in good condition. The fact that both recorders were found was what ultimately led to the final report and the details that I just told you.

So what conclusions did the final report draw? Why did this horrible accident really happen? Well, the fact is that no component of the plane failed, but the Pitot tubes and their heating elements were temporarily overwhelmed by external factors that were likely outside of what they had been certified for. This then caused a loss of reliable airspeed and that, in turn, caused a cascade of temporary system degradations, warnings and autopilot disconnections that caused serious shock to the operating crew. The crew was expected to apply the procedure for unreliable airspeed, but never did. Instead, the pilot's initial pitching reaction at the controls led to a continued worsening of the power situation that was lately discovered by pilot monitoring.

Subsequent corrections were too small, leading the aircraft to exit the flight performance envelope and enter a sustained stall. This fact was not understood by the pilot, most likely due to lack of training but possibly also because he may have thought that the warning was erroneous since it did not match his mental model of what was happening. In any case, these stall conditions were never discussed or verbalized by the pilots during the entire sequence. The failure to identify that position meant that the crew never initiated any countermeasures against it. This accident shocked the aviation industry and led to several safety recommendations, including the formation of a formal crash prevention and recovery techniques training module for all pilots.

That included the effects of high-altitude flight and stall recoveries. It also gave rise to new memory elements for flights with unreliable airspeeds that now focus on disengaging the automatics and setting the correct pitch and power immediately to keep the airplane flying safely before starting to troubleshoot. The investigation also recommended the insertion of a camera in the cockpit to display all instrument indications, as well as several other recommendations regarding pitot probes, black boxes, the Airbus stall warning system, as well as improvements in search and the rescue. Several other recommendations were also made, but the most important lessons that we, the pilots, must learn What we must take away from this accident is that we really need to understand the airplane we fly.

That means understanding all the systems, what happens if those systems degrade, and also how the aircraft might react differently when flying at high altitude or low altitude. And lastly, never forget to continue flying the plane. If a shock occurs, remember that tone and power will keep you safe. Keep the plane flying. And only after that try to find out what really happened. As long as the plane is flying, there will be plenty of time to fix the problem. As always, this story was based on the final accident report, but due to its complexity and the many human factors involved, I have also used several other sources and I especially want to mention Understanding Air France 447 by Bill Palmer, which was instrumental in expanding the understanding of my Airbus 330 system.

I highly recommend it. Now watch this video below or enjoy this playlist. Please consider subscribing to the channel if you think I've earned it, and if you'd like to support the work we do here, consider becoming part of my amazing Patreon team or buy yourself some merch. Have an absolutely fantastic day wherever you are and I'll see you next time. Bye bye.