ATR 72 Crash: Understanding the Flat Spin Disaster - Expert Analysis

[@quazz79]

As a former ATR72 captain, I've found the preliminary report a very hard read. I have no explanation as to why they continued the flight after the first Airframe de-ice fault. Diverting back to the departure airport was clearly the best course of action (and decision I have had to make numerous times). But to continue the flight and ignore multiple ice detector indications and all the APM warnings with seemingly no corrective action is gross negligence at best and criminal at worst. They had many opportunities to avoid this accident. It is something that should have never occurred.

[@NathalieV-c6c]

I cannot understand why the pilots would blatantly ignore such warnings when their own lives were at stake. Maybe we should be more careful than to use expressions such as 'gross negligence' and 'criminal' when referring to them. We still don't know why this happened nor do we know what the ambience in the company is.

[@isabelamuniz463]

@@NathalieV-c6c Yes, indeed, their lives were at stake. I believe they were negligent in this regard, and the pilot did not declare an emergency because, once on the ground, he would have to refrain from flying until an investigation was completed and he was... afraid? I wonder if he understood the gravity of the mistake he was making, if this had not been done before, and if he was overwhelmed with work, since he was not supposed to be flying that day. Many signs were ignored, signs that appeared even before departure.

[@Ficon]

My best guess is bad training, normalization of deviance, and get-there-itis. They likely thought they had time as they were expecting ATC descent into destination at any moment.

[@user-xu5vl5th9n]

How many times were they confronted with similar situations previously and how did they react? How many times were they assessed in training for similar situations and how did they react? The problem with abnormal situations is they are abnormal so there may be little practice.

[@Ficon]

@@user-xu5vl5th9n Normalization of deviance happens when you get confronted with a situation, do wrong things, and nothing bad happens. Given the local weather, they are likely dealing with icing all the time and have learned to ignore it rather than deal with emergency descents, etc. Proper stall procedures are part of basic flight training and upset training in the simulator. That they did not reflexively do the right things points to deeply flawed airline training and I think the airline needs to be shut down by the regulators until their training has been fully reviewed.

[@zigo7]

Dear Commander Magnar, As a retired airline captain here in Brazil after 38 years of flight, I would like to comment on the ATR accident. Regarding weather conditions, we are currently having a strong influence of "El Niño" that has formed atypical cold fronts in southern Brazil. The Voepass took off and flew inside a cold front that lowered the temperature a lot where it passed. Severe ice conditions advice was issued by meterological bases. Another observation I make is in relation to the captain who was young in age and perhaps in the function. I believe that they did not have specific training in the severe ice conditions, as it is a rare condition to find in Brazil, so we see that they did not take the actions they should or that the check list mandated. Not even the "memory items" were made. Another observation is to ask for a "mayday" or "Pan Pan" notice to the ACC and go down immediately. I believe there was resistance to issuing the emergency notice and having to respond later to your boss the reason. (Yes, there is such a thing here in small companies). Even knowing they had an inoperative defrost item. I reinforce that it is my opinion. I was an instructor and examiner in all planes that I flew and I witnessed this resistance to take certain actions because he thought he would be diminishing himself as a captain, unfortunately. Congratulations on the video and we will wait for the conclusion of CENIPA.

[@quackgarage9551]

Perfeitas observações. Complacência MATA. Tô cagando se a chefia vai ficar bravinha ou achar ruim, eu sou o cara que desliga o motor bom numa pane de bimotor em decolagem pra acabar com a assimetria se necessário, provavelmente o avião não vai ser recuperável, mas pelo menos eu e os passageiros vamos sair vivos.

[@luiscarlosdefrancamorais9569]

We don't have a el niinho in this time of the year

[@Musikur]

From the first moment of stall until they enter the flat spin is terrifyingly quick. If you ever needed an example as to why you should take early action and not ever let it get to the stall in the first place, this is it.

[@EstorilEm]

True, but this obviously wasn't a "normal" stall - the aircraft gave a number of warnings before the stall to avoid the situation. In a clean state (at least as far as contamination goes) the ATR stalls like a fairly normal plane from what I've heard. Throw in this many variables and all bets are off. It was definitely terrifying to watch, BUT it's also VERY difficult to get into the situation in the first place. It's almost like they were trying - or just complete ignorance thinking the plane would always save them no matter what they did (or didn't).

[@hugorneto]

From the Degraded Performance warning to the actual stall, 1 minute and 30 seconds have passed and the aircraft lost something like 20 knots in that time. It's so sad to see that, with the information that we have, this was apparently a totally avoidable tragedy.

[@NicolaW72]

@@hugorneto Indeed, exactly.

[@Izmael1310]

@@hugorneto pilots probably very occupied by this situation (lac of proper training) or they just did not pay attention to the situation unfolding in front of them.

[@rhr-p7w]

This channel is by far the most respectful and technically accurate data source from ATR related topics. All the other channels regarding the Brazilian crash are absolute rubbish.

[@marcosboaventura1977]

agreed.

[@CanalQuadrodeGiz]

can't agree more!

[@numismatajaime7974]

You are wrong. "Branco Aviação" did a excellent report about the crash.

[@oakromulo]

not all but most indeed!

[@geraldo209]

The blancolirio channel is a very high level of professionalism as well

[@gabrielguimaraes1425]

Magnar your approach to this topic and the insights you give are absolutely necessary and respectful. As a pilot and a Brazilian I can say that we appreciate your work and you are doing a very important work for the safety of all of us that are flying. Thank you so much

[@nakfan]

👍

[@NicolaW72]

👍

[@joaogabrielavanci138]

A respectfull explanation! Thank you! I’m a private pilot who lives at the city where de flight have departed, Cascavel, Brazil. We here are very sad about this tragic event. God Bless you and always give you safe flights and motivation to continue sharing with us your experience and great technical proficiency.

[@nakfan]

👍

[@NicolaW72]

👍

[@andreasheierre8915]

WOW - just WOW! Thank you very much for this exceptionally clear and detailed explanation, Magnar! This is the best aviation-video I have seen for ages! My very best regards! Andreas (Norwegian Air Traffic Controller)

[@TraderAmol]

can someone tell me what exactly happened?

[@jjstevenz]

Magnar I can hear the pain in your voice because you know you coulda helped those people live. God bless you my man. I felt it too. Pilots PLEASE listen to what hes saying and learn!

[@Hpeter695]

These are the best videos by far about this mishap. Full respect to everyone involved, no attention seeking, just the facts explained with true professionalism. Thank you!

[@beakman1966]

For those blaming the pilots, remember that the Preliminary Report shows the "what". The "whys" will be shown in about a year, in the Final Report. Never jump into conclusions.

[@marcg1686]

Correct, we shouldn't jump to conclusions. And this video doesn't do that. We know with reasonable accuracy what happened. It remains to be determined why it happened. Blame, responsibility and culpability lie close together.

[@TeaMohn]

This explanation is so masterclass even for anybody who has no flight license. Huge thanks sir 🙇🏻

[@ajs41]

I'm visiting this channel after a recommendation from Mentour Pilot.

[@jimlthor]

This dude is great. I don't even know how i found him years ago, and I barely knew what an ATR was, but I'm still here.

[@lmb5529bml]

Absolutely amazing and informative video as always Magnar. Cheers!

[@FlywithMagnar]

Thank you! Cheers!

[@ikefork2606]

Brilliant discussion on stalls, and stall prevention. A master-class level breakdown of this terrible tragedy. You are indeed an expert aviator, educator, and seasoned pilot. Thanks.

[@criminy_]

Thank you for covering this, Magnar. I'm curious to see the details that will be in the final report, but your analysis of the preliminary report and simulation video gives a lot of insight into this tragic accident. I hope that the aviation community can take the lessons of this accident to heart so that we can avoid these dangerous situations and loss of life in the future.

[@RicardoGonzalez-oy4oj]

You are the best!! You are an unbelievable source of knowledge for a father whose boy is going to flight school next year. God bless you!! Keep sharing your unique experiences.

[@fabianoabate]

The most clear and concise explanation ever about this sad accident . Thank you!

[@eugeniomj]

Very professional and respectful in your remarks. Thanks Captain! Greetings from Brazil!

[@donators1]

Certainly the pilots made the mistake by not take actions to avoid the accident! Many lifes were lost! Trainning and Trainning until failure never be an option! Best explanation I have seen from someone expert!

[@jorgeiman6678]

Dear Magnar : Your videos , specially the ones related to severe icing encounter, are crystal clear. All Turboprop training program - and I mean all, not only commercial - should include videos like yours to create clear conscience of the "famous" situational awareness : 1.- Observation, 2.- Orientation, 3.- Decision, and 4.- Action Thanks !

[@A2TS-n9l]

Amazing to see such technical explanation without any kind of speculations. It's a sad reason but thanks for this kind of approach.

[@nakfan]

These two videos on the accident are very, very good. So clear and well put together. For the first time I under stood the tragic flat spin. Thank you. BR, Per (Denmark) PS: This type of plane is used on our local routes…

[@justicemaake684]

Magnar is one pilot I can listen to anytime regarding ATR plane as he's speaking from experience not from opinion like the rest and he doesn't sensationalize matters for views

[@LeaoDN]

Wow... Great and clear explanation about how the flat spin was kinetically achieved! Impressive. Thanks!

[@spetzspas]

That speed tape at 0 is the stuff of nightmares

[@dougfraser77]

I wonder if this was a case of "normalisation of deviance" by the pilots. Perhaps they had flown close to the edge of the envelope in icing many times, and got away with it. With time that led to complacency, and a blasé attitude to icing. And eventually this caught up with them.

[@carlosfernandes8919]

Concordo com Voce, Eu acho que já eram acostumados, e sempre deu certo... Mas , Eu acredito que tenha algum agravante. tipo.... De Repente, uma Das Bolsas que arrebenta o Gelo , pode não ter funcionado... imagina , fazendo Gelo somente em uma das asas, no vooo inteiro... e Eles sem perceber...

[@jeffconstantine4341]

Normalization of deviance. 100% on the nose. It’s the only explanation for them not reacting to all the alerts.

[@MrCaiobrz]

I agree, both from complacency and maybe lack of training

[@carlosfernandes8919]

@@MrCaiobrz Eram muito treinados, aqui no Brasil o treino é severo...

[@deounb]

@@carlosfernandes8919 I think it happened due to the culture of tomorrow (manãna) of most of Latin America, that is doing nothing and hope for the best outcome. Complacency

[@lucianobrito1276]

Hello Mr. Magnar, I want to thank you for your explanation and since the beginning of the information about the crash I already believed that ice was one of the main causes of the accident, I live below the flight route and that day it was very cold and rainy. Unfortunately, it was a combination of factors that led to this tragic accident.

[@Skinny-me]

Incredible walkthrough! Your knowledge and methodical teaching skills are as always Top Notch! Thanks Magnar & Fly Safe!

[@lubeme1439]

Gracias Senor Magnar. Lupe Bernal. Mexico.

[@JavierBrent]

Ahi Chihuahua.. !

[@aileron5655]

Hello Magnar, thank you for this great explanation

[@secondskins-nl]

Really interesting review Magnar. It must be hard for families of all who died to hear this was something that didn't need to happen. I hope it will at least be a nice reminder for other ATR pilots who might get into the same circumstances. I can also imagine that severe icing might be more rare in Brasil but that's why you train such rare events in a simulator. Sad for everyone involved.

[@NicolaW72]

Not only a nice reminder - a lesson to learn.

[@baldock320]

Spectacular analysis

[@bonnoteriiteporouarai7097]

People are still blaming ATR, it’s so frustrating. The plane did it works, the crew’s lack of action is the reason why the plane went down.

[@NicolaW72]

Exactly.

[@JavierBrent]

I saw a post saying 30 ATR's crashed already for flying in ice. Liars love to lie. Only 5 in 40 years stalled with ice. All due bad pilots that stalled it even with all warnings to keep speed up. They were stupid to not do their job.

[@bonnoteriiteporouarai7097]

@@JavierBrent yes exactly

[@stefansekulic7903]

​@@JavierBrentYeh the aircraft is generally safe, it's only had 13 accidents in 35 years of service,but almost half were due to icing, not sure how that compares to other turboprops in its class.

[@FlyingSurprise]

This is one of my favorite flight related channel along with Mentour Pilot, Captain Joe and others. Excellent job!

[@mateusrodrigues1236]

Hi, Magnar! Again, thanks for your necessary and respectful videos that are amazing contributions to aviation safety. Considering the asymmetric thrust, I do believe they where flying with some rudder trim to compensate, I do believe it also may be a factor on the "fast" loss of performance, considering that it increase the drag. Would be nice to have your thoughts on it.

[@uninvited298]

It seems to me that regardless of the cause, in several recent disasters in which an aircraft reaches a stall situation, the pilots seem to act only instinctively trying to pull the nose of the plane up instead of lowering the nose to regain speed and control, as they should. The same reaction can be seen in the tragic accident of Air France Flight 447, where the pilots in command ignored the instruments and stall alarms and kept pulling the stick up until they completely lost control. First there is some confusion in understanding the situation, and then failure to follow the proper procedure. The same thing seems to have happened in this disaster. Of course, looking at it from the outside seems like an obvious flaw, as it's just a matter of the laws of physics, but it's intriguing how common this loss of situational awareness is in these disasters. They seem to completely forget that they have altitude, that they have room for maneuver, that they can ask traffic control to clear the way. Very intriguing.

[@foobarf8766]

The stabilizer stalls before wing afaik, to pull up is checking for authority. They had none. It's not the mistake.

[@TheOnePurpleGuy]

@@foobarf8766 Of course pulling up is a mistake in a stalling situation, especially for planes with a T-Tail. Although I'm not a pilot I have to wonder, why would you ever check for authority in a stall? To check whether or not you are already in a deep stall, and pretty screwed? AFAIK, any stall recovery procedure requires you to reduce the AOA immediately, and not check if your elevator is already stalled, and why would you? Also, I believe that most simulators are poor at simulating the stall characteristics of a plane, icing will only make it more unpredictable, as the accumulation changes the stall characteristics. The stabilizer usually does not stall before the wing does, because their airfoils are different, and symmetrical airfoils, the ones used for the horizontal stab, have a higher critical AOA. If the stab had stalled first, we would have expected a sharp pitch-down moment, but we just don't see that here. It looks like the wing is stalled first and they enter the spin by applying right rudder. After that the nose pitches down, as is expected in a spin.

[@foobarf8766]

@@TheOnePurpleGuy idk of course, all speculation but the pull up could also be dual control conflict. I don't know. But the stab size is smaller. The lift is function of area, if I had ATR in XP id try it (does laminar flow the new msfs still puts 172 into 'brick' stalls). Speaking of the c172 it has massive stab because it's a trainer. The stab is not main lift device ofc but does have +ve (not zero). CoG would determine pitch when losing it and it would be aft afaik? So pilot's intertial frame is experiencing

[@foobarf8766]

@@TheOnePurpleGuy pitch up not down if it stalled first because normal force

[@jamescollier3]

yes!! exactly!!! just like AF: aviator vs pilot. some just do what they were taught without the big picture, and some are actually aware of the big picture

[@andybrockbank3027]

Excellent commentary and explanation with a succint statement on actions to recover from an impending stall.

[@randercintra]

Great analysis of this accident.

[@dronemonkey2038]

This is a very good analysis…Thank you……What about precession? The power goes on…..the props turn faster giving a right hand precessing yawing moment…Further increasing AoA on the Left wing and it departs to the left side.

[@arb6591]

Thanks!

[@FlywithMagnar]

Thank you!

[@weldonyoung1013]

Magnar you nailed it. Correcting with the rudder just increased the angle of attack and stalled the airfoil of the main wing. Too bad the flight crew even got that far - so as to enter severe icing.

[@pgilb70]

Just escape from icing immediately. Especially when anti icing is hving issues! As blancorolio reminds us- a wing can stall at any speed but only one critical AOA. ...Nose down, escape!

[@nickloh912]

I wonder if severe icing being an emergency is not being emphasized enough. Certainly the statement "no aircraft types certified before 2015 are approved for flight in severe icing" tells us that a severe icing encounter requires action to escape, since it is a situation the aircraft is not designed or demonstrated to handle. But on the other hand, I can imagine a pilot being reluctant to declare mayday because of ice, especially if other aircraft nearby seem to be getting away with it. Even having read the report on American Eagle 4184, "severe icing is an emergency" was a bit of an eyebrow-raiser. It makes sense, of course, but it does emphasize the urgency of the situation more than "flight in severe icing has not been demonstrated, nor is it required under the certification rules" or something similar...

[@marcg1686]

You make some good points. Having watched many videos relating to flying into severe icing I would like to see ATR change the APM 'Increase Speed' advisory to an emphatic 'Increase Speed NOW!'

[@NicolaW72]

@@marcg1686 Exactly.

[@pamshewan9181]

Thank you!

[@MarceloTrindade1]

Great presentation! Thanks! I wonder what would be the purpose of an alarm that "you should never see in flight because you have to act before you see it"... Shouldn't it be changed from "Increase Speed" to "You're dead"?

[@msromike123]

Well in this case they had about 1 minut from that indicator until the stick shaker activated, much more until stall entry. As he said, "you still have time to save the day."

[@NicolaW72]

Probably better to: "Act or you´re dead!"

[@antoniosanna8]

Great explanation Captain, as usual 🎩 🔝

[@Nico-kz9bv]

I really dont understand that much as Im not from Aviation, but this incident fascinates me so that I want to understand things. Thanks for the upload!

[@fightingfulcrum8108]

Thank you very much for your very informative video and explanation Captain. Cheers!

[@christiangavrila]

This is how you intentionally initiate a spin for aerobatics: pull to stall and use rudder. Then instead of power idle, use full power like they did and you can get a flat spin. For non pilots it may seem that this is happening too fast, but it's not. In smaller general aviation planes this will be much faster and you should react immediately, like steering a car when about to leave your lane. Everything happened in the clouds with no visual references and when they saw the land they were already in a flat spin. You can see the pilots not only making mistakes but reacting late, behind what is happening, so most probably their upset recovery training was poor. It looks similar to AF447 except that in this case the plane worked with no issues.

[@juanmanuelgomezmartinez3795]

For flat spin you need CG aft that is normal in combat airplanes (unstable) but abnormal in transport airplanes

[@Karibanu]

In this case the plane had a de-icing fault. With how it departed I'm a little surprised how long it kept flying, to be honest. There must have been some horrific ice buildup, I wonder if they really had much tail authority at all & I'd hate to guess what the effective trim really was just before it departed. At face value this does seem a case of acknowledging the situation & blindly carrying on anyway - it will be interesting to see how much training they had around icing conditions & whether that really impressed just how much of a danger it is.

[@theearthwillshake8706]

Wouldn't it be quite different to AF447 aside from the stall aspect? If I recall, the Air France flight crew spent minutes misunderstanding everything that was going on and pitching the nose up into a stall - caused by faulty AoA due to frozen pitot tubes, whereas it appears that this crew just failed to react in time to the ice and low speed and stalled.

[@Karibanu]

@@theearthwillshake8706 AF447 reacted( poorly ) when the right thing to do was basically do nothing - these guys seem to have done basically nothing when the right thing to do would be fly somewhere else.

[@NicolaW72]

@@Karibanu Indeed, exactly.

[@130sogair]

Great explanation, Captain

[@geraldo209]

Thanks a lot captain. From a Brazilian new sub you got. and 737 driver.

[@wololo10]

737 pilot, melhora seu ingles amigao

[@76notam]

hello , j'adore vos videos , bonne description et premiers commentaires techniques de ce malheureux crash ...une question me vient à l'esprit : " pourquoi les pilotes ont ils laissé l'avion dans cette compsosante de givrage sévère et ne sont pas intervenus selon la procédure de chez ATR en pareil cas ? " l'enquête le dira certainement ..merci de vos informations claires et précises ..fly safe captain 😉✈🤙

[@MatthijsvanDuin]

Obviously this raises some questions about pilot training and/or airline safety culture (did they not know how to respond to the performance warnings, or were they habituated to ignore them?), but I think it's also fair to ask whether it should really be considered normal for an autopilot to happily trim a plane all the way into stick shaker and then unceremoniously dump the controls of the mistrimmed aircraft into the hands of a startled pilot. Is it really too much to ask for an autopilot to prioritize keeping the airplane within its flight envelope over whatever was requested through the mode control panel? Surely it could have noticed (or even predicted) the AoA approaching stick shaker and reduced bank angle to avoid it? I think ideally an autopilot (in the absence of a serious fault) should also avoid sudden disconnects and instead would *warn* the pilot if it is has trouble executing the commands from the pilot (e.g. due to approaching control limits), but continue to fly the plane to the best of its ability to allow the pilot to assess the situation and physically take the controls _before_ disconnecting the autopilot.

[@keithwalker6892]

@@MatthijsvanDuin After reading all the comments I still stay with my first one. This aircraft is too dangerous in ice and the pilots who fly them don’t seem to understand how sensitive they are by the number of fatal crashes. I worked for many years as a Certification Engineer Head of Structures and if still working would and will urge Transport Canada to withdraw the Type Approval of these aircraft

[@rafael.retratos]

então, no dia aqui no Brasil varios pilotos relataram gelo severo incomum no Brasil, inclusive avioes pousaram e aeroportos da minha cidade que ´é pequena, a aviação civil no Brasil é uma das melhores do mundo com baixa taxa de acidentes, mas neste caso alem do gelo severo, o piloto ser novo , um pioto foi chamado para cobrir outro de ultima hora, a investigação deve falar sobre fadiga, descanso, treinamento etc......

[@Coops777]

Thanks Magnar. This is by far the best explanation I've seen. Thankyou. I must ask the question of why the pilots chose to navigate and keep ATC happy in preference to aviating and flying the aircraft correctly. While it is very probable the pilots were unfamiliar with icing symptoms and procedures, they ignored speed and stall warnings to the point that when turning right (in accordance with ATC), the left wing stalled. The aircraft was still saveable at this point!!!!!!! It appears to me that that these pilots have not had sufficient, regular and relevant emergency training. The scenario of stalling/avoiding stalling was completely foreign to them. I'm putting my money on the airline management getting the blame for this accident.

[@LeaoDN]

I agree to everything you said. And I would like also to stress that the stall happened while the Auto-pilot was engaged. This does not take out or put any additional blame to the pilots, but the fact is that autopilot was engaged, initiated the turn as pilots have programmed it to do, and then the aircraft stalled. I hope this fact will also be studied by the investigators. It would not be hard to make the autopiloting system aware of the high probability of the aircraft having a degraded performance capability due to ice accumulation.

[@Coops777]

@LeaoDN Absolutely. I totally agree with you. And, another thing, according to another video I watched (might have been Magnar or Blancolirio) auto pilot IS NOT TO BE USED in severe icing conditions, according to the POH (or airliner equivalent) after the previous icing accidents in this type. It will be interesting to have this verified. But it makes a lot of sense. I am amazed at the lack of awareness and even skill by the pilots. The flight instruments themselves were a give-away (let alone the annunciator panel) . Increasing nose attitude (AOA) showing on the attitude indicator screen and decreasing airspeed whilst engine rpm and torque remained relatively stable at cruise levels (in the early stages) were sure signs of an impending stall. Then the aeroplane expressed its first obvious bit of body language and the left wing stalled during a right turn. Surely this would have been enough for the pilots to wake up and realise something was very wrong and take action to manually disengage the auto pilot and push the nose down. This is stuff they would have learned in GA school flying a warrior or a cessna.

[@LuizDahoraavida]

Thank you for the video!

[@markcardwell]

Thank you

[@McGyverPilot]

Yikes! Thank you for this more thorough analysis. On next video please show what pilots should have done after making any fatal mistakes. Remember, it might not be you as the PIC, it might be someone not educated or observing these cardinal rules, and therefore we might need to know what final actions provide any remote chance of flat spin recovery, no matter how impossible doing so may seem.

[@dougfraser77]

@FlywithMagnar assuming you find yourself in an ATR in a flat spin, would differential thrust on the engines help to arrest it? How much time do you need to recover?

[@jetdriver]

A flat spin in an ATR is likely unrecoverable.

[@MegaMadmechanic]

Not a pilot, but I have read that the weight of the engines in a twin-engine flat spin cause so much centrifugal force, it is very unlikely you would be able to recover

[@Jamenator1]

@@MegaMadmechanic This theory is correct, however it's more applicable to light twins. The main issue for a transport aircraft is likely the weight of passengers/cargo in the fuselage and fuel in the wings. The engines (excluding props) on the ATR are under 500kg each and are only about 4 meters from the CoG. Pax/cargo load can be around 5,000kg, and half of that load is probably further from the CoG than the engines are. Plus fuel in the wings which can account for another several thousand kgs, also quite far from the CoG. At the end of the day, the engines probably contribute a tiny amount of annular momentum compared to the fuel and pax/cargo load

[@wesleyhurd3574]

It seems to me that the autopilot should automatically disconnect when the degraded performance indicator is illuminated. The autopilot continuing to attempt control of a compromised aircraft only made matters worse. The pilots were likely confused by the unexpected feeling of the flight controls at the point in time when the autopilot was finally disconnected. Clearly, the continued flight in icing conditions is likely to be the biggest contributing factor to the crash. However, in order to prevent a similar crash from ever occurring again, it would be wise to find any point at which the chain of events could be stopped. In the interest of human factors engineering, the manufacturer should also consider making the “increase speed” indicator red in color to reinforce the concept of an impending catastrophic failure.

[@marcg1686]

I'm no pilot but I disagree. Disengaging the AP has to be a conscious decision. The AP may be masking unequal aileron loads. That's why it has to be disengaged prior to entering a region of severe icing. Manually controlling the plane will allow the pilot to better sense imbalances that indicate ice accretion.

[@josephaltieri1017]

excellent analysis,

[@danabergman4052]

Thank you captain very informative

[@regionalflyer]

@11:35 I flew the old 212/212A. I seem to remember a pneumatic system that would force the condition levers to 100 OVRD automatically under certain circumstances. It's been 14 years since ive flown it so I could be wrong, or mixing it up with one of the other many types ive flown 😂. One thing I do remember is I hated that POS lol.

[@shakeyspizza01]

Thank You Captain

[@keepthefaith-theconcerttribute]

As a former pilot this is very hard to stomach as all the warning signs were there and yet it got to a point where it never had to get too. Swiss cheese model in full effect here sadly.

[@michaelkaliski7651]

The aircraft is trimmed by the autopilot to a full nose up setting and the wing and control surface performance is degraded by ice accumulation. Then the autopilot hands back control to the pilots, who have perhaps ten seconds at best to recognise the situation, trim nose down and gain control of the aircraft with degraded controls. With the element of confusion and disorientation, it was never going to happen. The airframe de-icing fault was undoubtably a distraction that contributed to the confusion when the autopilot disconnected.

[@aipravoce]

I live in the same region of this accident. There was a great temperature change with strong winds. But it is a relatively common phenomena. There was a 15 or 20 celius degrees variation very fast.

[@NicolaW72]

Thank you very much for this very enlightening Explanation! It really shocked me because it seems to be that the Pilots made even more serious and eventually deadly mistakes than I recognized until now, allthough I followed all the information about the crash very closely so far. So far as I can see would be only explicable with a serious lack of training - and that would mean: It has something to do with the Airline. It seems to be that both Pilots were totally unprepared for how to act in severe-icing-conditions and how to recover from a stall.

[@diogonascimento5087]

Melhor explicação até hoje foi melhor que a que o cenipa fez

[@LeaoDN]

O CENIPA foi transparente e forneceu os parâmetros de forma detalhada. Impressionante o que as pessoas competentes como o Magnar podem fazer com o auxílio da internet nos dias atuais.

[@Saltlick11]

Great report

[@750cruzerpilot]

I wonder if the g forces combined with startle prevented flap 15 from being selected

[@MarkxTube]

Thanks for the higher res animation. Do we now have stick and rudder imput data as well? Could it be that the horizontal stabilizer was already shaddowed by the main wings before entering the flat spin? ie the pilots/pusher could no longer initiate a dive? And why is there no more voice data towards the end if we still have flight data. Is there a lag in recording or just not released yet?

[@foobarf8766]

Caveat I'm not an aviator but I think so, yes, the stab stalls before wing does in most traditional ampenages. You lose elevator authority in sims at least, have more roll. Except blended body elevons they stall different

[@gibuja]

For me, it looks like a poor CRM. There was no voice to be recorded.

[@misarthim6538]

I eagerly await the report on this, it's hard to parse what kind of reasoning would make the crew fly an aircraft with INOP deice into known severe icing conditions. I also hope that some part will be dedicated to a fact that - yet again - the crew decided to fight the stick pusher. Likely it's an instinctive reaction but there must be a reason why this happens in not insignificant numbers in similar accidents. (just to be clear - I'm not blaming the pilots, we simply don't know why it happened. We need facts. I was merely observing that it's quite shocking and I'm interested how it happened).

[@uwemobil8847]

As a passenger, that I am, this is hard to watch and to accept. They had the time and the possibilities to get out of that situation.

[@numismatajaime7974]

Magnar is a gentleman to not say this, but it is clear that the pilots caused the crash.

[@dougfraser77]

The underlying question is WHY - were they under-trained, were they complacent? What is the safety culture at their airline? Were they under pressure to stick to schedule?

[@Studz-gt]

@@dougfraser77 So far only speculations. They could be undertrained for ice conditions considering that s a rare situation you encounter in brazil. Or maybe they were suffering from fatigue. I also heard from a brazilian pilot that in Brazil if you declare an emergency then you have to go through a lot of bureocracy with the government once youre on the ground to explain what happened etc. So pilots feel more aprehensive about declaring an emergency. I know sounds messed up, but again these are just speculations. Final report might help clear things up.

[@foobarf8766]

Remember the swiss cheese model. Of course the pilots crashed it, but that's the obvious thing, what made them so complacent?

[@marcg1686]

@@Studz-gt 'They could be undertrained for ice conditions considering that s a rare situation you encounter in brazil.' Nope. In one of Magnar's older videos he describes flying into a region of severe icing. In Thailand.

[@jamescollier3]

yes. you can point at a lot of things, but this is like AF stall into the ocean. There are aviators that fly their private planes, maybe race or compete. Then there are pilots that have learned just enough to pass and fly. This is true in almost any career

[@brandyballoon]

I'm dumbfounded. The pilot inputs were correct... if they were trying to put it into a spin! I have a horrible feeling this was one of those situations where the FO knew what to do but was either too scared to speak up or was overridden by the captain. I find it hard to believe that both of the pilots forgot the basics.

[@Taketimeout3]

Objective analysis without subjective opinion. Just the facts and their relevance.

[@LeaoDN]

The initiated curve also introduced a yaw due o asymmetrical drag, right? How much is hard to know because the surfaces were badly contaminated with ice.

[@pfsantos007]

Maybe I missed it, but where does the "cruise speed low" get its info from? Do you enter planned cruise speed in the 'computer' (FMS or MCDU equivalent for ATR)?

[@FlywithMagnar]

The APM computer receives data from many sensors and computes the theoretical cruise speed. Then, it compares it with the actual cruise speed.

[@carlosfernandes8919]

Sou do Brasil , não sou Piloto , mas Voce Explica muito bem... Como é impressionante, que já cairam muitos Avioes , por motivos Semelhantes... 2 Boings 737-Max, Airbus A-330 Air France e Vários ATR-72

[@Antonio-lt1sp]

Os 737 MAX não foram stall.

[@stephenj4937]

The other crashes happened very quickly (less than a minute for the Max crashes, and 1-2 minutes for Air France). This accident happened very slowly and with lots of warnings that were completely ignored.

[@MrCaiobrz]

The MAX crashes where based software after a AOA sensor failure, the AF 447 where bad training in reacting to a temporary pitot sensor fault, and some ATR due to icing. Not at all similar

[@carlosfernandes8919]

@@Antonio-lt1sp Verdade, alta velocidade e Nariz embaixo, Por falha Eletrônica, mas é muito Rápido pra se ferrar...nisso é tudo igual...

[@carlosfernandes8919]

@@MrCaiobrz Parecido na Rapidez do Acidente...e Falta de conciencia da situação...

[@msromike123]

Is not this exact scenario not trained in the simulator? If not why? and if so, did these pilots do not receive it?

[@wewillrockyou1986]

Very strange decision-making by the pilots in response to the icing and supposed anti ice faults. I am curious on what basis they make the assertion that the system is faulty, are there any indicators for it in the cockpit? And also, very strange control response by the pilots during the onset of the stall, I am reminded of AF447 and the MCAS crashes that pilot response to unusual control situations is sometimes or even often absolutely not in line with what the regulations and aircraft certification expects.

[@normannutbar424]

It has been shown time and time again that even very, very experienced crew, when shit hits the fan and they are faced with the terror of a deep stall, do not push!!

[@Izmael1310]

In one of the videos you said that to recover stall you have to follow the "PARE" and one step is the "rudder oposite". Now you said they should not use the rudder.

[@FlywithMagnar]

PARE is for spin recovery. For stall recovery, push the nose down and keep the other controls neutral until you have a safe speed.

[@ecstazyrm]

Excellent thanks very much. If you were there what would you have done differently Magnar?

[@dougfraser77]

I'm sure he would have reacted much earlier to the initial issues with the de-icing system -- certainly not flying into icing with the de-icing switched off -- and never got into the situation in the first place.

[@MrCaiobrz]

He would have followed his correct training. Training clearly lacking here.

[@tedwalford7615]

What should have been the first, best response of the flight crew? At the first deicer-fault alert, should they have dropped out of the icing flight level and returned to base?

[@HarryFortyTwo]

The elephant in the room: why did they use rudder? There is the potential they had hinge moment reversal and could not get wings level with the ailerons bc of that. Flying so long in icing with AP on (which is exactly what is not recommended) and then causing a stall is the perfect recipe for hinge moment reversal. and that could also explain why they did not get the nose down, trying to get wings level first (which of course is not the most important thing in that situation).

[@ajnsouza]

I wonder if it was wings freezing in the very dense clouds? In the south and southeast of Brazil at this time of year there is a lot of rain and sometimes frost. Unlike Voepass when it comes to the Northeast even in winter, here it's strong winds and rain, but it's very warm

[@luigilainom1123]

I'm not a pilot but if your plane is stalling, rudder and ailerons neutral and pitch down

[@grahamsalmons2027]

What was the terrain elevation over which they were flying?

[@georgeedgeworth915]

Could it have been the yaw damper have moved the rudder to the right?

[@AdairGalloJunior]

Maybe they had the MAX-MCAS accidents in the back of their minds as a reason to “fight” the stick pusher. 11:49

[@NikolaiUA]

I believe the yellow AoA line should be below, not above the green pitch line in the graphic.

[@davidkavanagh189]

Is there any possibility the horizontal stab is stalled too? It seems hard to believe they would have the strength to oppose both the stick pusher and the aircraft's natural tendency to want to pitch down into the now downward vertical flight path.

[@FlywithMagnar]

The only instances of horizontal stabilizer stall occurred when full flaps was deployed during the approach. The nose of the aircraft wound then drop straight down.

[@NikolaiUA]

If the aeroplane was already stalled, why were there two "bursts" of violent climb as per the VSI? I think it reached about +15K fpm. How can a stalled plane climb?

[@danielbeards6259]

Magnar, could the yaw damper be responsible for the rudder kick right before the spin? How does ATR AP react when stalling? Does it disconnect both AP and YD or does it remove aileron and elevator servos, but keep YD (and therefore rudder servoactuator) alive?

[@FlywithMagnar]

The yaw damper provides yaw damping and turn coordination. It cannot move the rudder to full deflection. The autopilot disconnects automatically when the stall warning is triggered. The yaw damper remains engaged. The yaw damper disengages when a force exceeding 300 N is applied on the rudder pedals.

[@wayne_kurr]

Does the ATR not have an auto throttle system? I note the power in the cruise sits at 82% N1 until the pilots attempt the stall recovery. This is devastating for the families, and equally terrifying that the stall recovery was handled so poorly - let alone the list of bad decisions and lack of airspeed awareness that got them into that position to start with.

[@FlywithMagnar]

Tuboprop engines have different indications that turbofans. NP is propeller rotation speed. It is 82% in cruise regardless of power. NP was increased to 100%, and during the final moments, the torque (TQ) was incrased.

[@wayne_kurr]

@@FlywithMagnar thanks for the reply. I’m a 737 driver. Why wouldn’t the AT system try and hold the target cruise airspeed? Edit - just realised my error in reading NP1 vice N1 - still, I would have thought the TQ would increase as IAS decreased.

[@FlywithMagnar]

No autothrottle in turboprops. The power lever is set in a fixed position called the "notch." A power managelement selector can be set to Take-Off, MCT, Climb and Cruise. The engine will produce that power. We do not cruise with reduced power. The autopilot can follow ALT hold or IAS hold. Not both.

[@wayne_kurr]

@@FlywithMagnar I see, thanks. Very unfortunate sequence of events.

[@zlm001]

Thanks.

[@pdfernandeza]

lack of training, take a look at the training background of those pilots, there will be the answer

[@NicolaW72]

And this answer will probably be located at the involved Airline.

[@BreakTubes]

Mentour Pilot channel brought me here. Now I´m hooked!

[@CarlosAlexandre-fl2ut]

I don't see how a "lack of training" will be avoided on the final report...

[@jetdriver]

Because it’s pretty obvious that they didn’t do what they were trained to do. Every airline pilot in the world is trained at a bare minimum to run checklists when they get a fault and they ignored multiple lights that should have required them to run a checklist. Not to mention the totally botched stall recovery. Though if they were trained in an old school fashion to never lose altitude in the recovery that might explain at least that much. Time will tell.

[@luigilainom1123]

@@jetdriver they weren't airmen, there's not excuse.