What's new

737 MAX Fiasco

user-74gear

New member
The tone of your channel seems to be more upbeat so I’m not sure if you’d like to tackle this.. but I’d love to hear your thoughts on the 737 MAX situation. For those unfamiliar, this plane had a defective angle of attack sensor that was feeding bad data to a system called MCAS, and the result was the plane sometimes going into a dive. Two planes crashed, more almost crashed, and a lot of people died.

I’ve heard arguments from people that know a lot more about aviation than me that the whole concept of the MCAS is a kind of abomination, where the plane can only fly under the guidance of a computer because of what some consider a flaw in its design.

Then there was a whole related scandal about the bad design being rooted in Boeing trying to avoid pilots having to get certified.

Boeing got fined billions of dollars over this so there’s obviously a consensus that they screwed up, but there are also some more nuanced issues here about aircraft design, the role of computers in aviation, certification, the complicity/negligence/incompetence of the FAA here, etc. It would be cool to get a video on it.

Here is some discussion on it from my own professional (IT) community:
 
Last edited:

Rick O'Sidhe

New member
There's quite a bit of disinformation and speculation out there that was driven by the media. At the end of the day, you have to ask yourself why a US-based carrier - who are the largest global operators of the MAX - didn't have a hull loss incident involving MCAS. The short answer is that their pilots likely followed the published procedure for runaway stab trim to the letter, and have practiced recovering from that condition in the simulator.

Was MCAS fatally flawed? Nope, but it had flaws. Did Boeing cheap out by not including multiple AoA sensors in the 'base model' MAX, which allow the avionics to generate an AoA Disagree Alert? Absolutely. Did the AoA Disagree Alert actually work as it should? Not really. Did Boeing do an inadequate job explaining MCAS automation to pilots? Again, yes. Were the airlines complicit in this, in order to minimize required training and avoid a new type certification requirement? Yup. Did Boeing do a good job of letting pilots know of which QRH procedures were affected (and how) by MCAS? Nope.

The last item, in my opinion, is what really confused the doomed pilots and led to the hull losses. Prior to the MAX, you could disconnect the autopilot from trim by flipping a single cutout switch (Stab Trim Auto Pilot). You then flip a second cutout switch to disengage electric trim motors (Stab Trim Main Elect). QRH non-normal Runaway stab trim procedure calls for flipping both to cutout for continuous runaway - some pilots, however, leave the electric trim motors engaged (if they determine that runaway trim isn't due to the motors) to retain trim switch function on the yoke. With the MAX, an unannounced change was that MCAS enablement is tied to both switches - both MUST be set to cutout to disable MCAS. When the pilots couldn't manually spin the trim wheel (presumably) due to aerodynamic loading, they flipped Stab Trim Main Elect back to normal, unintentionally re-engaging MCAS. Had they unloaded the elevator first using any of a variety of methods, they would have been able to manually adjust stab trim, recovering control of the aircraft. Instead, they diverged from the QRH procedure and ran out of both time and altitude, which ultimately had fatal consequences for hundreds of souls.

At the end of the day, all (sadly) share in fault - the manufacturer, the operators and the pilots. I do put most of the blame squarely on Boeing and the airlines, though. Commercial imperatives should *never* trump life safety. As with all automation systems, simplicity for the operator (flight crew) shifts complexity behind the curtains. With aircraft automation, you simply can't make major adds or changes to systems and not communicate to and train pilots on those systems and changes. Today's aircraft are so highly automated that it's impossible for a pilot to know everything about everything, but that's why manufacturers and operators publish the QRH and pilots train on it.
 
Last edited:

Rick O'Sidhe

New member
My feeling is that over the decades the 737 QRH has probably got so thick with additions that the 'Q' part is probably historical.
There's been a bit of that, of course. But there is also a QRH "memory items" list of procedures. It's a short list which addresses conditions that present an immediate threat to safety of flight. Runaway stab trim is one of the QRH memory item procedures. While the procedure has changed since the days of the -100, it hasn't changed very much. It's not an easy task for a pilot with, say, 8K hours in type, to retrain their memory to account for frequent revisions, so any changes are highly scrutinized before (if) receiving regulatory approval from the FAA.
 

Plague

Active member
It's not an easy task for a pilot with, say, 8K hours in type, to retrain their memory to account for frequent revisions
I do think that there was (is?) rather too much dependence/expectation of the pilots picking up any errors the machines make. If they are busy or just plain tired they can miss clues, so should be helped when possible.


any changes are highly scrutinized before (if) receiving regulatory approval from the FAA.
Er, yeah ... That kinda leads to the other little problem.


Runaway stab trim is one of the QRH memory item procedures.
I get the impression that the MCAS+AoA failure doesn't present quite the same. Eg a stuck trim button would result in a continuous run of the wheels which is quite noticeable, whereas the MCAS just keeps applying small increments over time which could be missed since the trim often does these movements anyway (in auto flight).


I do put most of the blame squarely on Boeing and the airlines
I put almost all on Boeing.
Who in their right mind builds a system with the authority to crash a commercial flight based on a single sensor input - and a sensor that is quite vulnerable as well.
I used to build control systems for factory automation and tried to avoid or mitigate that sort of situation, even though there was no (obvious) threat to life.
 

Zeede

Active member
Staff member
I put almost all on Boeing.
Who in their right mind builds a system with the authority to crash a commercial flight based on a single sensor input - and a sensor that is quite vulnerable as well.
I used to build control systems for factory automation and tried to avoid or mitigate that sort of situation, even though there was no (obvious) threat to life.
Redundancy is key to aircraft safety. This is why even a lowly Cessna 150 has two sets of magnetos. Boeing should never have delivered a plane without a backup of some sort for such a critical component.
 

Roichi

Active member
I would add some blame on the FAA and congress.
The FAA for insufficient regulatory oversight and congress for insufficient funding.
Especially the FAA had a big role in that as it had in other accidents in the past. Part of that problem is that the regulations which are made by congress are heavily influenced by the aircraft manufacturers and their interests in profit. Thanks to some well placed campaign money and threats of closing factories in electoral districts. Or open them in others. Corruption at this level is some part of the problem.
Other is that the FAA has an unhealthy connection to the manufactureres. They know each other too well. People transfer from one to the other and the manufacturers have political influence on the authorities. That leads to insufficient oversight where menufacturers can certify their own systems on behalf of the FAA.
 

MsHighAltitude

Active member
Staff member
How come nobody has blamed Wall Street? :) Boeing probably wouldn't have done what they did if it weren't for earnings pressure.
 

Plague

Active member
How come nobody has blamed Wall Street? :) Boeing probably wouldn't have done what they did if it weren't for earnings pressure.
Ultimately everything comes down to money/profit. Proper governance should keep things sensible ... but that's getting political :cautious:
 

Penfold101

New member
The whole system stems originally from Boeing trying to be cheap and rush a larger 737 varient to the market to compete with the A320 NEO after they saw it's success. Instead of designing a new airframe (which would've taken a decade or so) they went with modifying the current 737. This is all well and good, except that this design is essentially the same as the original -100 model. Over the years it's been stretched and re-engined, without much change, if any, to the height it sits off the ground. This is determined by the length of the undercarriage legs, and that is limited to the space in the fuselage, which when all taken together gives you a limit on what you can change.

As the original model had turbojets, the engines were very skinny and nowhere near the ground. With the fitment of turbofans, the engines became much closer and as such had to sit quite far forward of the leading edge - in front of the wing, rather than under it. You can see how much of an issue it is when you look at the intake - the lower part has always been flat rather than round to accomodate enough ground clearence.

With the MAX, they fitted such large diameter engines that required placement even further forward of the wing that it massively upset the stability of the whole design, which isn't suprising given how old it was. The solution was MCAS - making the pilot feel like he was flying a normal 737 when the computers were constantly keeping it stable. This would require almost no extra training and be very attractive for the airlines. From here the points mentioned above about the AoA sensors, the checklists and the lack of training are all relevant.

It's also worth mentioning that the gear had to have system installed to enable it to shorten upon retraction and thus still fit in the bay. Upon selection, it would be compressed to a shorter length before being retracted. If this doesn't highlight the inadequecies of the design and re-use of ideas from the 60's...
 
Last edited:
Top