The 737MAX. What is wrong?

After the recent crash of the Ethiopian 737MAX8 I started to receive a lot of questions like "what's going wrong with the MAX", "what do you, Denis, think about this accident, shall I change my tickets to Airbus", "how to live now", and many others.

First, guys, I want to stress, that the investigation has just begun and at this moment there is a lack of official information.

However, this accident looks very similar to the first crash of the same model which happened half a year ago in Indonesia with a Lion Air 737MAX8. Both airplanes crashed just in a few minutes after airborne and before impact they were integral and the vertical speed and the air speed were so high, that almost no big parts of the airplanes remained, just small fractures.

Besides this, a pilot of the Ethiopian 737 declared emergency and reported that they were having problems with control of the airplane. And after a while they found a part of the stabilizer screw and said it was in the full nose down position. 

Ok, todays' discussions and theories are based on the facts, that the Ethiopian airplane has crashed just after takeoff, crashed into pieces, and this looks very similar to the first accident in Indonesia.

* * *

So, what's going wrong with the MAX? Let's review the history. I want to underline that everything I am going to say is based on what can be found from public sources and the preliminary report of Indonesian accident, which is also in the Internet.

After the crash in Indonesia the 737MAX pilots all over the world got surprising news that there was a new system installed on the MAX, a Maneuvering Characteristics Augmentation System, or an MCAS! We didn't know about the MCAS before the crash in Indonesia! This is the truth.

The MCAS enhances pitch stability characteristics of the airplane when flying with flaps UP and at higher angles of attack The MCAS counteracts the bad influence of new big engines on the aerodynamics. According to what we know now the MCAS works only if the aircraft is being hand-flown. If a specific high angle of attack is sensed the MCAS starts a slow motion of the stabilizer to nose-down with maximum portion of 2.5 degrees which takes approximately 10 seconds.

During this, if the pilot trims the airplane by the manual electric switch located on any yoke column, the MCAS stops its work. But if after 5 seconds conditions for its activation still exist the MCAS starts another incremental stabilizer motion for another 10 seconds and another 2.5 degrees. And at the end the MCAS can move the stabilizer to a full nose-down position.

The MCAS will also stop its work if the angle of attack has decreased to a predetermined value.

The natural reaction of pilots when they heard about the MCAS was: "What a fuck? It would be good to know about such a system!"

Nevertheless, there was no description of the MCAS in the flight crew operations manual, FCOM. There was an only word MCAS. Guess where? It was in the abbreviations table.

It is now a fact that in Indonesia the MCAS activated because one of the sensors of angle of attack was producing wrong data – much higher numbers than actual! It looks unbelievable, but Boeing didn't implement any cross-check of parameters, and this erroneous data was accepted by the MCAS and it started to command a nose-down motion of the stabilizer thus forcing the airplane to dive and loose altitude.

In turn, this caused a lot of problems to the crew which had been already overloaded by another outcomes of the faulty sensor – the yoke was shaking as the system got a terribly high angle of attack, and at the same time and because of the erroneous data a difference in airspeed and altitude indications appeared between primary flight displays as the airplane systems use angle of attack data to apply corrections to the barometric speed and altitude.

Besides that, it was a very early departure. I think, this is obvious that the crew was sleepy at this moment and their readiness to troubles was significantly reduced.

What is very important in the story of the Indonesian accident, that in the last flight before the fatal one, everything happened the same – the pilots of that flight got all the same failures. But luckily, they managed to survive, by just using standard actions which… and I want to stress this… have been written in the manuals since first 737s of Jurassic generation appeared in the skies. They did Runaway Stabilizer memory items.

The pilots recognized the uncontrollable movement of the stabilizer and disconnected the stab trim cutout switches as prescribed by the book.

By the way, recently we got news that there was a jump-seater, another qualified pilot, who noticed the movement of the stabilizer trim whell and suggested to do something with it. And this saved the aircraft.

So, switching it off stopped the electric motor and thus stopped the work of the MCAS. They landed the aircraft safely by using manual trim of the stabilizer.

Are they heroes? Yes, they are... Partially.

After finishing the flight, they made two entries in the tech log: "Altitude disagree" and "FEEL DIFFERENTIAL PRESSURE light illuminates". They wrote nothing about a stick shaker, nothing about airspeed disagreement, and nothing about their fantastic fight with the stabilizer.

No report – no action! This is a rule in aviation. So, the technician did the trouble shooting of the entries, they did some tests and let the airplane go into its last flight. The reason – a faulty angle of attack sensor which was installed before this flight – wasn't found. And we all know the result.

And now, after half a year another crash happened with the same aircraft model in very similar circumstances. Is it a tendency? It looks like it is.

* * *

In one of Russian major aviation forums I was asked like: "Hmm. It looks like there should not be any difficulties with the MCAS even if it is faulty?"

This question was based on what Boeing had said about the MCAS.

"Normal electric trim control on the yoke can stop the MCAS driven stabilizer movement, however MCAS will activate again within 5 seconds after the switches are released if the angle of attack is still sensed too high. Pilots can deactivate MCAS and automated control of the stabilizer trim with the cutout switches and can hand-crank the trim wheels on each side of the throttle quadrant for manual trim."

That's all. Looks pretty easy, isn’t it?

Let's come back to the history. A few days after the first 737MAX crashed, Boeing managed to issue a safety bulletin in which they described what to expect in case of an erroneous angle of attack data. 

Just look at this list:

And they prescribed action, which looked pretty simple: what to do in case of uncommanded stabilizer trim – just do memory items which every 737 pilot must know by heart – but are they really simple? I will come back to this trap later in the part II of this video.

All right, it really looks smoothly on the paper. Just do what you've been expected to do since first 737s came into operation. And… I gave a following answer on this question:

This is what I have been thinking about since getting the first news about the sneaky operation of the MCAS. The sequence of how things may be going in the situation looks like the following:

1.The airplane takes off and just after airborne pilots  get a stick shacker on the captains side and in a while they get a malfunction, unreliable airspeed and altitude. Pilots start doing memory items from the QRH NNC, which requires them to disengage the autopilot and the autothrottle, switch off the flight directors and set certain thrust and pitch:

If flaps are up: 4 degrees and 75 percent N1

If flaps not up: 10 and 80.

2.If the pilot holds the pitch, then his attention should be concentrated on the primary flight display. Otherwise he cannot control the pitch. And because he has to set the pitch precisely, he must trim the airplane in order to remove any forces on the yoke to zero, so that if he removes his hands from the yoke the airplane continues to maintain the pitch with that thrust.

3.As the pilot must trim the airplane, he is using the switch on his yoke, which – as we now know – disconnects the MCAS for another 5 seconds. Boeing says and the FDR data of the Indonesian flights prove that the main electric trim from the yoke switches works properly or at least is supposed to work properly. By using the switch, the pilot can reverse the MCAS negative inputs and trim the airplane normally.

4.Let's suppose that 5 seconds have elapsed and the MCAS starts doing its sneaky job again – slowly commands another nose-down. If the pilot is still flying the airplane – I mean, he is maintaining the pitch – he obviously starts to feel the work of the MCAS as appearing of pulling forces on the yoke which he must apply to maintain desired pitch. By the way, I'm not talking about the noise which the stab trim wheel produces during its rotation.

5.A natural action of a pilot if he feels constant forces on the yoke which impede him to fly the aircraft precisely is to trim the aircraft in the direction of the forces he needs to apply. Thus, using the yoke switch will be his next action naturally. And this will stop MCAS for another 5 seconds.

6.So, this loop will continue to happen again and again until pilots finally recognize that something goes wrong, and they've got conditions for a Runaway Stabilizer NNC.

This is how things happened in the first flight of Lion Air which ended safely even though the crew didn't know about the MCAS. And this really look smoothly on the paper, and after Boeing has issued their Bulletin all 737MAX pilots are supposed to understand the problem and know what to do in case they get similar conditions during takeoff. Again, I am planning to come back to this trap later in another video.

Why this scenario didn't work in the consequent flight of the Lion Air 737MAX? The flight recorder shows that for 6 minutes the crew was fighting with the MCAS! For a period of 6 minutes the crew was periodically applying a nose-up trim and the MCAS was applying a nose-down trim! Looks too long, what could have prevented the crew to understand the situation and apply needed actions like the previous crew did?

* * *

I have the only idea – stress. A very big and sudden stress.

Look at the situation from a different angle. It was an early departure; the flight crew obviously wasn't in their best condition. And just after the airborne they got:

First. A stick shaker on the captain's side. This can be a stall! A big danger! Instinctively pilots would like to push the yoke in this case after having so many simulator trainings in the past. This is already a stress and a pressure.

Second. Unusual forces on the yoke which can appear because of two things: the incorrect wok of the elevator feel system. This also can happen because of the activation of a stall protection mode.

Third. Warnings that Airspeed and altitude are unreliable.

Now you must do a QRH checklist and switch off the flight directors. And this is a big trouble in modern aviation. I published a lot of articles about the problem of degradation of basic flying skills because of too-much reliance on automation nowadays.

And not only me – a lot of pilots all over the world claim the same issue! Pilots of big jets are rarely flying departures and approaches manually and even if they are doing so, they use the flight director guidance most of the times. Which is, believe me, not the same as you are flying the airplane without the flight directors!

A lot of pilots fail their simulator assessments during initial screenings because they can't show their skills of flying the aircraft using raw data! The FAA in the recent past issued a SAFO, which prescribed the US operators to implement such automation use policies which would allow and encourage pilots to keep their proficiency of manual flight, to maintain basic skills of flying, which is a must to every pilot, not only for weekend flyers of Cessna.

A year ago, I wrote an article about how to cope with Airspeed Unreliable on a 737. And I said:

"Pay your attention, please, that memory items say you to disconnect both FD switches. Yes, I fully understand your happiness, because finally you've got a chance to fly the airplane without flight directors.  But, would I be a passenger of your flight, I don't think I'd be happy… if your airline policy doesn't allow you to train such flying on a regular basis in normal situations. I think, we understand each other."

Fourth. G-loads. As the aircraft is flying up and down, it is obvious that all people onboard will experience very bad feelings and pilots are not an exclusion. Negative and positive Gs greatly affect our consciousness and ability to understand what is going on.

So, all this shit happens just after airborne. When nothing happens to you for many days, months, years – and this is true, rare pilots face with real problems nowadays – every human is prone to become too much relaxed.

We cannot be stretched like a spring all the times being always ready for a shit. During simulators we are aces as we know that shit will obviously happen today and even more, sometimes we know exactly what shit will happen during this session. And our spring becomes less stretched.

In the real life, shit happens suddenly.

* * *

My thoughts above can look like I'm saying that these were pilots who crashed the airplanes. No. Every accident happens because of many factors. It is a chain – just break it somewhere and shit will get out.

What happened to the AOA sensor and why it has happened? Could the Lion Air ilots of the first flight describe the problems in the tech log? Could the pilots of the last flight manage to apply QRH actions? Finally, could Boeing engineers exclude this scenario by just producing an MCAS which would cross-check two AOA sensors or recognize malfunctions by other means?

This is an aviation. We, pilots, are at the top of the system. And I'm just talking from pilots' side. Not all the times airplanes are perfect, not all the times the weather is nice and not all the airports have long dry runways with ILS or GLS installed. Not all the times everything go as expected. Very rarely, but lot of shit can happen at the same time. Should a pilot be prepared for problems? This is what passengers expect from us. This is what the industry expects from us.

You know, if I find myself in a situation in which I must do procedures which I'm expected to know by heart and I fail to do them, I wouldn't be happy if somebody says that my fault happened just because of the aircraft. It will be my fault anyway, even though there was a lot of factors why I had failed. This is my position – we are sitting in the flight decks not only for posting nice pictures in the Instagram. We have been put there to cope with issues – because of aircraft systems, bad weather, bad crew planning, bad ATC, unruly passengers and so on. That's why we are being asked first if something happens. No complaints from my side.

Behind that, of course, all factors should be investigated and corrected. Risks of forcing pilots to be heroes must be minimized! This is how the system should work.

Safe flights!