# I Can’t Believe They Did THIS! | USAir Flight 5050 ## Метаданные - **Канал:** Mentour Pilot - **YouTube:** https://www.youtube.com/watch?v=pxGiKrdBiIo - **Дата:** 30.05.2026 - **Длительность:** 42:40 - **Просмотры:** 467,875 - **Источник:** https://ekstraktznaniy.ru/video/52700 ## Описание Use the code "pilot" and this link 👉 https://incogni.com/pilot to get a whopping 60% off the Annual Incogni plan! ----------------------------------------------------- On the rainy night of September 20th, 1989, a Boeing 737-400 thundered down LaGuardia's runway with nothing mechanically wrong with it, yet within seconds it would shoot off the end of the runway and into the dark waters of Bowery Bay. So how did this happen? Let's find out. ----------------------------------------------------- If you want to support the work I do on the channel, join my Patreon crew and get awesome perks and help me move the channel forward! 👉🏻 https://www.patreon.com/mentourpilot 737 Course: 👉🏻 Visit https://www.virtual737course.com to learn how to fly a 737 like Petter in our virtual 737 SIM course Make sure to use code "mentournow" for 20% off! ✈️ Our Connections: 👉🏻 Exclusive Mentour Merch: https://mentourpilot-shop.fourthwall.com/ 👉🏻 Our other channel: youtube.com/@MentourNow 👉🏻 Our Podcast: ## Транскрипт ### Segment 1 (00:00 - 05:00) [] This aircraft is thundering down the runway through rain and wind with its two pilots desperately trying to control it. But nothing is actually wrong with the aircraft itself. So what's about to cost this absolute nightmare is something else entirely. By the evening of September 20th, 1989, the entire New York air traffic system was under a lot of strain due to weather, congestion, and sheer traffic volume. Low clouds and intermittent rain had been present over the region for most of the day, forcing aircraft into longer routings and holding patterns, whilst controllers tried desperately to feed arrivals into one of the busiest airspaces in the world. At Laguardi airport, the delays had begun striking up since early afternoon. And that kind of operational pressure was not unusual in New York, but LaGuardia amplified it in ways few airports could. You see, unlike the enormous hubs like the nearby JFK, LaGuardia had been built into the edge of the city itself, boxed in by water and urban development with little room to expand and therefore almost no margin for recovery once traffic began falling behind schedule. This meant that aircraft arriving from Washington, Boston, Pittsburgh, Baltimore, Norfolk, and dozens of other East Coast routes were therefore packed tightly together from early morning until late at night. And by September of 1989, the airport was handling nearly 60,000 passengers per day. Caught in the middle of all of that on this evening was a nearly new boring 737400 in a gleaming silver US Air livery and with registration November 416 uniform Sierra. This 737 with its advanced CFM56 engines was one of the newest variants of Boeing's increasingly dominant short hall airliner and the aircraft itself had accumulated only about 2,235 flight hours since delivery less than a year earlier. and it was in a great overall condition. It had arrived with its crew at Laguardia earlier that night after a delayed flight from Baltimore and was at this point sitting at gate 15 while the ops controllers tried to make up their minds about exactly what to do next. The airline operating it was at this time in the middle of a transition of its own. Only weeks earlier, the plane's original owner, Pedmont Airlines, had formally merged into US Air as part of what was at the time the largest airline merger in American history. [snorts] But large mergers like that normally doesn't happen seamlessly from an operational standpoint. And that's because procedures, training programs, cockpit habits, dispatch systems, and company cultures all had to be blended together. While the airline continued operating at full schedule, so pilots who technically now worked for US Air had often been trained under Pedmont systems and procedures, and many crews therefore still carried habits and assumptions from their old previous airline. — [snorts] — In the left seat on this particular night was a 36-year-old captain. He was an Air Force Reserve pilot who also flew the Lockit C130 Hercules outside of his airline work. And he had accumulated 5,525 total hours, including roughly 2,625 hours on the 737, but only about 140 of those had been flown as pilot in command after having upgraded to captain only 2 months earlier. His training records generally described his performance as average. He had passed his similar training without any real issues, but that training had focused mostly on catastrophic events, not smaller series of compounding issues, which is worth remembering. Like many airline pilots at the time, he had also never received any formal crew resource management training, which is training that places heavy emphasis on things like communications and tightly structured crew coordination during both normal and abnormal situations. Beside him, he was joined by a 29year-old first officer who was still almost entirely new to airline jet operations. He had accumulated 3,287 hours of total experience, mostly on smaller aircraft and commuter turbo prompts, but he had only 8. 2 hours in the actual 737 outside of simulator training and initial operating experience. And this was actually his very first day on the line after his training. So in practical terms, he was likely still transitioning psychologically from his highly scripted environment of training into the far less predictable rhythm of actual live performance. And as I've explained many times before on this channel, that would have made him especially dependent on following precise procedures as that's really all ### Segment 2 (05:00 - 10:00) [5:00] new pilots have to hold on to. Now, he had actually been hired by Pedmont Airlines and had then been folded into a US Air employment when the merger completed, but given his limited experience, that culture shock would likely have been smaller for him than for many others. Now, both of these pilots had been trained and qualified to the standards of the day. But for those of you paying attention, there was still a little bit of a problem with this particular pairing. Combining a newly upgraded captain with a brand new first officer was of course not ideal, especially when combined with an operation already strained by weather delays, schedule disruptions, and constant changes to the evening's plans. And with the lack of proper CRM training, the effects of all of those factors would be even more pronounced, as we will soon see. Originally, the crew's next assignment after arriving from Baltimore had been a routine flight over to Norolk, Virginia, but that flight was then cancelled. And shortly afterwards, they were reassigned to an empty repositioning flight over to Charlotte, North Carolina, where the aircraft would then be needed for another flight. But then that plan also changed. The company now decided that some passengers who had been stranded by cancellations and delays at Laguardia would be boarded onto this aircraft instead for an unscheduled extra flight over to Charlotte operating as US Air flight 5050. Now these repeated changes visibly irritated the captain who was becoming increasingly concerned that any additional delays could push the crew towards their duty time limits before this night was over. So whilst gate agents worked to reshuffle passengers and update the manifest, he repeatedly left the cockpit to speak with dispatchers and operations personnel about these scheduling decisions. And that meant that he left the first officer alone in the flight deck supervising the boarding and preparing for departure since he was going to be pilot flying for this coming leg. And as he was doing that, the first officer was also joined by a Panama captain who was dead heading, meaning that he basically was just tagging along as a passenger without a confirmed ticket. And when that captain entered the cockpit, well, then he sat down on the jump seat behind the captain's chair and made himself comfortable, possibly with one foot on the center pedestal. Meanwhile, other US Air employees kept shuffling in and out of the cockpit, delivering weather information and updated flight plans to the first officer, who then took some of that paperwork and his approach charts and temporarily placed it on the center pedestal. Eventually, the captain came back and the final pre-flight preparation was started, but curiously, there was no formal emergency briefing done, nor any discussions about how this coming takeoff would be completed. And that was even though this captain knew that this would be the first officer's very first takeoff after his line training and in quite challenging weather and conditions. In fact, it was later found that no emergency briefings had been done during the full 9 hours that these two pilots had worked together, which is pretty extraordinary. At one point after the boarding had been finally completed, US Air then tried to get the captain to reopen the boarding door again to let 15 additional passengers on. But the already frustrated captain just said that the crew would be running late and possibly into duty time limitations as it was. So he therefore refused to take those extra passengers. A decision that possibly ended up saving lives. Now, almost every accident we investigate on this channel involves enormous amounts of data being collected, stored, shared, and passed around without the people involved ever realizing. Things like maintenance records, communication logs, and more. And as it turns out, your personal data is not so different, which is why it is more important than ever to protect your online privacy. And that is exactly what today's sponsor, Incogn, will help you to do. Incogn will scan the internet for sketchy data broker sites who collect things like your phone numbers, address, social security numbers, and family information and then sells that on to the highest bidder, which could be a marketing agency, insurance company, or even criminals. When Incogn finds your data on one of these sites, they will issue an immediate demand for the data to be deleted. And legally, these sites then have to comply. Now, sure, you could do that yourself as well, but it will take a lot of time and work to do, and you will then need to repeat it every few months when your data has been harvested again. But when you sign up with Incogn, they will constantly scan and keep removing your data without you having to lift a ### Segment 3 (10:00 - 15:00) [10:00] finger. And you can then check their progress through their dashboard whenever you want. Plus, you will get regular status reports sent to you. And you won't believe just how much data they have removed from me since I signed up. Incognate now also has a custom removals feature where you can send an exact link to your exposed data if you find it and their privacy experts will do the rest which is pretty cool. Right now you will get a fantastic 60% discount if you scan this QR code or go to incogn. com/pilot. Just remember to use the code pilot at checkout. Thank you incogn for sponsoring this video. Now let's get back to the story. Anyway, after the original boarding, the plane had 57 passengers and six crew on board, bringing the total up to 63. But by the time that the aircraft finally pushed back from the gate at time 2252 local, investigators would later conclude that the accident sequence had probably already started. You see, as the passengers settled into their seats and the last bit of paperwork was organized around the cockpit, one small control, sitting quietly on the center pedestal, had likely already been moved into a position that would seriously change the behavior of the airplane. And that control was the rudder trim switch. During normal flight, pilots use rudder inputs through the yaw dumper to keep the aircraft coordinated during turns. But we also use it to compensate for crosswinds during takeoffs and landings or to counter asymmetric thrust after a potential engine failure. But because holding a constant rudder pressure with your feet for long periods of time would obviously quickly become exhausting, the aircraft also includes a trim system that allows the pilots to apply a small continuous rudder input without physically maintaining pressure on the pedals themselves. Under normal conditions, especially on takeoff in relatively light winds, the rudder trim would typically remain centered. And in fact, most pilots rarely touches it at all outside of abnormal situations. I have only used it for very small corrections in crews and after engine failures in the simulator. Now on the 737300 and 400 series, the rudder trim switch was positioned at the very back of the center pedestal between the pilots and within easy reach of anyone sitting in the cockpit or moving throughout it. The switch was at the time blade shaped and in order to activate it, the switch only needed to be held slightly away from the center in order to then activate an electric motor that would reposition the roer pedals and the position of the roer. Now, the investigators would later be very clear that they couldn't know when the rudder trim moved out of position since the flight data recorder wasn't active during the turnaround, but the only conclusion that they could draw was that sometime during the boarding, the switch had been activated and then held counterclockwise until it had moved into the nearly full left trim position. Now, the switch would have needed to be held to the left continuously for around 30 seconds for that to be possible. And two of the theories put forward about how that could have happened were that the first officer had activated it when he showed a binder of paperwork in between his right seat and the center console, or that the jumpse seating Panama captain briefly had rested his foot on the console and inadvertently activated the switch when he did so. Sadly, we'll never know exactly how it happened, but the important thing is that none of the pilots had noticed it. Now, that oversight becomes a little bit hard to understand when you look at what the rudder trim actually does to a 737 while it is still on the ground. You see, it is linked directly to the rudder itself by means of the built-in hydraulics, but it also is linked to the nose wheel steering during taxi. This means that this left rudder trim would have displaced the rudder pedals by depressing the left one in this case. And if that wasn't noticed during the pre-flight, which it should, it would have created a subtle tendency for the aircraft to drift left during taxi, just like if one of the pilots were constantly pushing on the left ro. So the captain who will be taxiing using his tiller would therefore have to compensate continuously to the right. And that's even before the actual takeoff would begin where the effect would become even more apparent. But I think subtle is an important word here. Taxiing a large airliner through Laguardia at night in the rain and the wind already involved constant steering corrections, breaking adjustments, radio calls, checklist work, and attention divided between taxiway traffic and changing departure sequencing. So maybe ### Segment 4 (15:00 - 20:00) [15:00] a slight pull to the left would not necessarily have felt that unusual particularly on wet pavement with the feedback through the controls was naturally reduced. Another factor was the wind here who if it was coming from one of the sides would have pushed the large fin of the back of the aircraft and therefore have created a similar need for compensation. And even though that wind would not have moved the rudder itself on the 737, we have to remember that this captain had flown the C130 Hercules before and it had a tendency for the rudder pedals to move with the wind. So it is possible that this captain with his limited command experience on the 737 just thought that this displacement was normal. Anyway, all the cockpit preparations were eventually completed without anything strange having been noticed. US Airflight 5050 soon requested clearance for push back from their stand and to start up their engines. Once that had been completed, they then started taxing out through the rain and wind towards runway 31 with the captain now likely already compensating for the abnormal trim condition without consciously noticing that he was doing it. Ahead of them, aircraft were lining up on the different taxiways waiting for departure. Though as flight 5050 continued working its way towards runway 31, they were at some point told to hold position briefly at one of the intersections, but that was missed by the crew who just continued taxing until the ground controller intervened and gave them an updated taxi route. Now that mistake would come to have absolutely no impact on what was soon about to happen. But from a pilot perspective, this could have been an early indication of a maybe somewhat suboptimal cockpit cooperation between these two pilots. You see, the taxi phase is a sterile phase of flight because it is so easy to make mistakes here and it therefore requires full concentration by both pilots. The captain is the one with the tiller and therefore always the one taxing on the ground on most Boeing 737s, but he will still rely heavily on the help of the first officer who will be following through on the taxi charge and speaking up if something seems wrong. Now, that might seem simple, but it's actually a pretty advanced phase seen from a crew coordination standpoint. So, the fact that they made a mistake here could be nothing, of course, but it could also be a sign of what was to come. In any case, inside of the cockpit, the pilots now continued preparing the aircraft for departure using the standard before takeoff checklist. And it was here that one of the most deceptively important lines in the entire accident sequence quietly passed by almost unnoticed. As part of the checklist, the first officer called out, "Stabilizer and trim. " And the captain quickly replied, "Set. " He then caught himself and added, "Stabilizer trim. I forgot the answer. Set for takeoff. " But even though he did that, something important had now been missed. Did you catch it? Well, the wording of that checklist item concealed a subtle weakness that investigators would later spend enormous time analyzing. You see, on the Boing 737, the word trim refers to multiple different systems, including the stabilizer trim and the rudder trim. But in normal airline operations, most pilots would often just associate the word trim at this stage with stabilizer trim because that was the setting actively calculated before every takeoff. The stabilizer trim directly affects the aircraft's pitch behavior during rotation and crews therefore routinely verifies its position against takeoff performance numbers before departure. But the rudder trim not so much. The only time that trim is touched is during the early pre-flight sequence where its function is tested and then returned back to neutral. But after that nothing. So here, while the checklist had been completed, there was no evidence that either pilot physically verified the roer trim position at all. They didn't skip the item. But by not actively checking all of the different trim settings, the checklist itself had simply failed its purpose. Now, this is different than when pilots rhyme the checklist, which is where a correct checklist response has been learned by heart and therefore responded even if that item isn't done. Now, in this case, the captain likely did check the stabilizer trim, but just failed to check the rudder, which I kind of understand since it was located way behind him, and he was taxing the aircraft. So, that really should have had its own checklist point, maybe verified by the first officer. In any case, the captain then continued by briefing the departure using the ### Segment 5 (20:00 - 25:00) [20:00] takeoff speeds which had been calculated for now this very light aircraft and the most crucial item in that briefing was the V1 speed which is the speed where a decision to either reject the takeoff or continue has to be made. Beyond V1 we pilots are committed to the takeoff since the performance calculations then shows that beyond that we might not have enough runway left to stop. So any failures noticed after V1 should be brought into the air and then dealt with later. The captain now told the first officer that for this flight V1 was set at 125 knots. Rotation speed would be 128 knots and the safe climb speed after the engine failures V2 would be 139 knots. But he didn't go into any further instructions about how to actually deal with a rejected takeoff. Instead, he was just happy with the first officer reciting what his initial maneuvering would be after takeoff. Outside of the cockpit windows, the runway lights shimmerred across wet pavement while other aircraft continued departing one after another into the darkness over the bay. Now, at the northwestern edge of LaGuardia, where they now were, runway 31 stretched just over 2,100 m out towards the water, and it ended in a short or run before dropping away into the bay below. Under normal circumstances, that runway was more than long enough for a lightly loaded bowing 737400 like this one. But that was depending on the pilots following a very strict set of circumstances, which was especially true if the aircraft would have to reject its takeoff for any reason. At time 2318, ground control finally instructed flight 5050 to taxi into position and hold and to switch over to the tower frequency. And 2 minutes later, the tower then cleared the aircraft for takeoff. As it rolled forward and onto the center line, the captain now tried to lighten the mood slightly by asking the first officer if he was ready for it. And the first officer answered with a nervous laugh, saying, "Here goes nothing. " But as he began moving the trust levers forward, another important link in his accident chain had actually already been set in motion. You see, the very last point on the before takeoff checklist had been out to break. And when the first officer had challenged that item, the captain had just answered is off. Now the autobreak had an RTO setting which would automatically engage maximum braking in case the thrust levels would be above approximately 90 knots. And that would obviously be very helpful in reducing the speed as soon as possible. And it would also help the operating pilot by removing some of the workload if he needed to concentrate on controlling the aircraft, for example, during the rejected takeoff. So by electing to not use it, the captain was now removing that important safety barrier. So why would he do that then? Well, that's harder to say, but widespread misunderstandings about the function of the auto breakback then existed. A lot of pilots thought that it could cause the passengers in the back to hurt themselves if it would suddenly activate at relatively low speed and others thought that they might need the option of differential braking to maintain directional control after an engine failure. Now, both of these sentiments were largely false and not based on facts. But in any case, the first officer had accepted the captain's decision, and he was now getting ready for his first ever takeoff after his training in this pretty lousy weather. So, after having spooled up the engines and allowed them to stabilize, he now needed to release the brakes and then just press the toga buttons in order to begin the takeoff roll. Like most modern airliners, the 737 used an automated takeoff mode designed to reduce workload. So by pressing those toga buttons or takeoff goaround switches, the auto tht throttle would automatically advanced both engines to the correct thrust setting calculated for the departure. But as this first officer now reached forward to engage the system, he accidentally pressed the out of throttle disconnect switch on the truss lever instead, which is a mistake many new pilots have made. Now, at first, neither of the pilots fully realized that this had happened. So, the first officer tried to push the toga switches again. But because the outer throttle had now already been disconnected, the thrust levers didn't move. Instead, the aircraft just continued accelerating slowly down the runway using what little thrust they actually had set. Now, the captain soon recognized that something had gone wrong, saying, "Okay, that's the wrong button pushed. " And he then volunteered to set the power for the first officer manually by advancing the tropless to ### Segment 6 (25:00 - 30:00) [25:00] what he thought was the correct setting. But investigators would later determine that neither engine actually reached the exact calculated takeoff thrust setting. And they also noticed that the left engine ended up with about 3% less thrust than the right. Now that was not enough to create a dangerous asymmetric thrust condition by itself, but it did add to the already left pulling tendency of the rudder trim. And this was soon about to get worse. As the aircraft now accelerated through the lower speed range slightly slower than normally, the effect of the left rudder trim started to become more and more clear. You see, at lower speeds, the directional control of the 737 is dominated mostly by nose wheel steering through either the rudder pedals or while we're taxing the tiller beside the captain's seats. But as the plane then accelerates, the air will start moving faster over the tail and eventually cause the rudder to become aerodynamically effective and therefore take authority over the plane's direction. And when that happens, the effectiveness of the nose wheel steering becomes proportionally less important. But this also means that the trim part of the rudder becomes more powerful. So in order to keep the aircraft accelerating straight down the center line, the first officer now needed to push more and more right. Now a very important factor in a two pilot cockpit is the division of duties. One pilot will be pilot flying and only he or she will be controlling the aircraft. If the pilot monitoring would start to interfere with this, especially if that ends up being the captain, confusion about who is in charge can very quickly build up. And you're soon about to see an excellent example of that. Because as the first officer was now trying to keep his aircraft straight, the captain was also making steering corrections using his tiller, probably believing that he was helping the first officer. But because he was doing that, neither of the pilots now fully understood the exact control inputs that the other pilot was making. And because the airplane was still technically controllable and accelerating straight ahead, the severity of the situation wouldn't have been immediately obvious. Also, crucially, neither pilot communicated what they were doing to the other. So, the aircraft just continued accelerating down the wet runway while small steering corrections gradually became larger ones. And when they passed 50 knots, the aircraft began drifting left by a noticeable amount. And the first officer responded to this by pressing harder on the rudder with the captain still correcting with the tiller. Now, I should mention here that the tiller should absolutely not be used for directional control at high speed. Using the tiller is like moving the handlebar on a bike. It moves the nose wheel a potentially very large amount and at higher speeds when the rudder is dominating the direction, this will subject the nose wheels to potentially enormous side forces, forces that it's just not certified for. So, as this aircraft now accelerated past approximately 62 knots, the stress on the nose wheel tires became too much and one of them just burst. This meant that inside of the cockpit, the pilots now heard a loud bang from beneath them. And as the damaged wheel assembly then continued vibrating against the runway, a deep rumbling noise also started vibrating up through the whole aircraft. Now, until that moment, the crew had unknowingly been dealing with a steering problem caused by the rudder trim. But once that tire burst and that rumble began, their mental model likely shifted towards the possibility of some sort of mechanical failure. The aircraft was pulling left. There had been a loud bang and something underneath the airplane now sounded badly wrong. So, it no longer felt like a subtle control issue. It felt more like something serious and acute. And at the same time, the steering situation continued to get worse. The faster the airplane accelerated, the more aerodynamic authority the mistrimmed rudder developed, and the harder the first officer had to fight in order to keep the aircraft aligned. And now came the line that would later define the entire accident sequence. The captain suddenly called out, "Got the steering. " Which was also what the cockpit voice recorder captured. But afterwards, neither pilot agreed on what had actually been heard. The captain later testified that he believed that he had said, "You've got the steering. " Meaning that the first officer should continue controlling the airplane. But the first officer believed that he had heard, "I've got the steering. " Meaning that the captain was taking over directional control himself. So for the next several crucial seconds, both pilots partially relinquished ### Segment 7 (30:00 - 35:00) [30:00] responsibility for controlling the aircraft at the same time. The first officer eased off some of his rudder input, believing that the captain had taken over, but the captain had never fully transitioned into controlling the takeoff himself. And as an obvious result of this, the aircraft now started veering sharply further to the left, accelerating towards the grass on the other side of the runway. Now the captain who remember still thought that he was pilot monitoring should have called out 80 knots to indicate that they were now in the high speed regime of the takeoff and later also V1. But because both him and the first officer were so concentrated on the directional control of the aircraft, none of those calls were ever made. And I cannot think of a better way to highlight just how crucial proper crew coordination and communication is. This is what happens when those two things completely breaks down. In any case, a moment later, the captain finally made a decision to reject the takeoff. But when he did that, the aircraft had already accelerated to about 130 knots, which was 5 knots faster than they computed V1 or decision speed. And the way that he announced this decision would then make this situation even worse. Now, I just want to pause here for a second to clarify what should have happened here. A ro trim that is pulling the plane hard to the left is not a reason to reject the takeoff near or past V1, but it would have been a reason to do so before passing 80 knots, at which point the effect of this issue was already clearly noticed. But that would have required communication. Once the aircraft was in the high-speed regime, meaning past 80 knots, the best thing to do would have been for the first officer to just continue controlling it with his rudder. The rudder is more than capable to override the effect of the trim. And then he should have brought the aircraft into the air to sort it out later on, which he could still have done even with that damaged nose gear. The captain could have also taken controls at any point if he would have properly called it out, but it would likely have been hard for him to get the feel for the aircraft and the rudder inputs needed. So, the best thing would have been for an early rejected takeoff instead. And had it been handled in any of those ways, the nose gear would have still remained intact and the aircrafts would have either stopped or taken off safely. But what now happened instead was that with the aircraft still pulling left and the damaged wheel rumbling violently under them, the captain called out, "Let's take it back then. " And he then pulled the trust levers all the way back to idle. But that was not the correct call out to use. He should have called out, "Reject," which is short and unambiguous and something that his colleague would have immediately understood. Now, as soon as he had taken this decision and these actions, the aircraft actually began tracking slightly straighter than it had seconds earlier. And that was because the slight differential thrust had now disappeared. And the captain had also started applying differential braking together with aggressive steering corrections to pull the airplane back towards the center line. But using differential braking also meant that he wasn't using the max braking that the out brake would have done. Remember, he had left that one off. And since he was now concentrating on getting the aircraft under control, this also meant that valuable seconds passed before full reverse truss was selected, which at the speed that they were now traveling of around 130 knots meant that they were eating up 70 m of runway per second. Now, a rejected takeoff must be done in a very precise sequence in order to minimize stopping distance. What you do is that you first close the thrust levers, disengage the outer throttle, then you use RTO or max manual braking as required. Manually raise the speed brake lever in order to get all of the weight down on the brakes and then you select full reverse thrust. Now this is generally done by the captain. So the first officer therefore automatically becomes pilot monitoring and should call out any deviations, speeds and runway remaining. But since this captain hadn't used the correct phrase, it is likely that it took a while for this brand new first officer to realize that they were in a rejected takeoff. So, he might not have realized that the captain wasn't doing the things in the correct order or that he was missing things. In any case, a few seconds into this maneuver, the first officer radioed out, "Usair 50/50 aborting. " And the tower controller just calmly responded, "Rogger, left turn at the end. But inside this cockpit, a cold realization had now dawned on these two pilots. The runway was almost gone ahead of them, and they were still at high speed. The captain continued breaking aggressively was also trying to hold the ### Segment 8 (35:00 - 40:00) [35:00] airplane near the center line, but the wet runway and their delayed deceleration had at this point already sealed the outcome. Someone in the cockpit now yelled out, "We're going off. We're going off. " And then the pavement just disappeared. US Airflight 5050 now shot through the end lights, dropped off the elevated runway deck, and then slammed into the wooden pier structure that was supporting the approach lighting system extending out into Bowery Bay. The impact was so violent that it tore the aircraft apart as it fell down towards the water below with the nose section separating first as the lower fuselage struck the pier structure beneath the runway. Moments later, the main fuselage fractured again near the wing roots, while debris, luggage, and sections of cabin interior erupted outwards into the darkness. The tail section finally broke partially free as the aircraft came to a rest in the shallow waters north of the runway, tilted awkwardly with portions of the fuselage submerged. But fortunately for those on board, the crash had still been survivable. Because of the relatively low impact speed with the water, there had been no explosions or postc crash fire. Instead, the deadliest damage came from something far more localized. You see, as the airplane struck the peer support beneath the runway lights, portions of the structure pushed upwards into the underside of the cabin around rows 21 and 22 and the cockpit, and that crushed sections of the floor upwards into the passenger compartment itself. Tragically, this ended up killing two people and injuring the captain. But those two fatalities would end up being the only ones. But for everyone else, this nightmare was still far from over. The cabin lights failed almost immediately as cold black water began pouring into portions of the aircraft through all of the various holes and ruptures. Overhead bins had burst open during the impact, scattering bags and debris across the floor. But the passengers were now struggling to orient themselves in the darkness. The flight attendants had only had seconds to grasp what was going on and the fact that the aircraft was not burning, but instead sinking. So they now realized that they needed to get everyone out as quickly as they could. They soon noticed that the forward left entry door couldn't be used because of water outside and portions of the fuselage around other exits had been damaged or distorted by the breakup sequence. So the overwing exit soon became the primary escape route for much of the people inside of the cabin. Whilst water was continuously flowing in, the ditching lines attached near the overwing exits were deployed correctly by survivors and then stretched outwards into the water to give people something to hold on to as they drifted away from the airplane. Now, because this flight had not been scheduled to be operated far away from the coastline, the seats did not have inflatable life jackets installed, which would have been helpful. Outside, the conditions were now miserable. The rain was continuing to pour down while jet fuel spread thin shimmering layers across the water around the wreckage where loads of passengers were now swimming. And inside the aircraft, the cabin crew relentlessly continued directing evacuations despite the flooding water and the near total darkness. Now the tower had seen the aircraft carine towards the end of the runway without taking off and had therefore activated the crash alarm before the plane had even hit the water and rescue vehicles had therefore quickly raced towards the departure end of runway 31 whilst boats from the port authority police and nearby emergency services pushed out into Bower Bay through the rain and darkness. But water access routes were awkward. Visibility was poor and survivors were scattered both inside and outside of the aircraft across a wide debris field extending beneath portions of the runway approach structure itself. So roughly 90 minutes passed after the crash until the captain and the rescue crews finally became convinced that nobody else remained trapped inside of the wreckage. And only then did they finally leave the wreckage of the shattered Boing 737 sitting silently beneath the runway lights. In the weeks and months that followed, investigators would slowly piece together a sequence that at first glance barely seemed to make any sense. The [snorts] aircraft itself had been mechanically healthy. The weather, whilst poor, was well within operational limits. The crew had not suffered an engine failure, a flight control failure, or any other catastrophic malfunction that would normally explain a high-speed rejected takeoff ending up beyond the runway. So instead, what they found was something far more uncomfortable. This had largely happened because of a stunning lack of communication and crew resource management, which had turned a relatively small issue with a misset trim setting into a fullblown disaster ### Segment 9 (40:00 - 42:00) [40:00] and that therefore led to some broad changes in the industry. Now, the most immediate physical change involved the Boeing 737 itself. Investigators became deeply concerned about the design and placement of the roer trim switch on the center pedestal, particularly because it could apparently be moved accidentally by a binder, a hand, or even someone resting their feet near the console. In response, Boeing modified the design of the switch and the surrounding pedestal geometry on the 737 fleet. That original blade shaped switch was replaced with a round control and additional guarding around the aft portion of the pedestal was added to reduce the chance of inadvertent movement. This accident also accelerated broader adaption of formal crew resource management training. At the time of the crash, neither pilot had received formal CRM instruction, something that sounds shocking today, but was still quite common in parts of the industry back in 1989. So during the following years, the concept of CRM became far more deeply embedded across US airline operations. And by 1994, the FAA formally required CRM training for airline crews, part of which required calling out speeds and V1 checkpoints during the takeoff rule. The investigators also criticized the pairing of a newly upgraded captain with an extremely inexperienced first officer. And whilst no outright prohibition of similar crew pairings emerged directly from the accident, the industry increasingly moved towards more conservative crew scheduling practices involving low experienced pilots. Now, several operational changes also grew directly out of the rejected takeoff itself. The board emphasized that the aircraft's outer brakes had not been armed despite the wet runway and relatively short departure distance contrary to manufacturer and company recommendations. So following this accident, the NTSB pushed for greater emphasis on maximum performance rejected takeoff procedures, including more explicit use of auto brakes when available and improved pilot training on stopping distance assumptions during high-speed rejects. Now, if you want to learn how real pilots today operate the Boeing 737, complete with procedures, explanations, and checklists from Ben and I, well, then please scan this QR code or go to virtual 737course. com. Anyone who is curious about aviation will enjoy this course. And if you're flying home simulators, it will truly transform the way you do it. I promise you. My name is Peter Hornfeld and you're watching Mentor Pilot. Have an absolutely fantastic day and I'll see you next time. Bye-bye.