LarryJ

For how long was the A330 fleet grounded after the AF447 accident?

Productivity

Plato90s

Is it possible that with the 737-Max design being more prone to higher-than-wanted AOA than previous models (hence the need for MCAS), problems with sensors like the ones involved in the LionAir crash causes more pronounced issues than past models?

Is it possible that the key driver is bad batch of sensors?

MSPeconomist

jrl767

LarryJ

The MCAS system is needed because the higher residual thrust of the new LEAP engines produces a light control feel at high angles-of-attack near a stall or in steep turns. The MCAS system provides a nose-down bias in those situation in order to produce a control feel that is consistent with the existing 737 NG models.

TWA884

Here's a Permalink that should work.

MSPeconomist

Plato90s

Yes, I got that from the existing coverage of what MCAS is supposed to do.

It's a software-compensation approach to the changes that Boeing introduced with the 737-Max in order to create better fuel efficiency/range.



In IT infrastructure, the 737-Max is what I'd call a fragile infrastructure. It's a system which requires active tuning/intervention in order to maintain stability. That is in contrast with robust systems which are deployed so that it delivers consistent performance without the need for as much fine-tuning.

In a robust environment, I'd expect there to be no active intervention 95%+ of the time. In a fragile environment, active intervention by the management software kicks in frequently.

Robust infrastructure very often turns fragile when users try to push the envelope too much - putting more workload than the system is really meant for. It's possible to continue to deliver the same performance/behavior via the active tuning, but it's not an ideal situation because the fragility means it's a lot easier to trigger a catastrophic sequence which crashes the entire system. That usually happens when the business wants to increase efficiency by squeezing more work out of the existing infrastructure footprint.



I'm not a pilot nor am I involved in aeronautics but the coverage thus far on the 737-Max suggests that Boeing has pushed the envelope too far on the 737-Max and made it a fragile system.

apodo77

Southwest CEO sent out a letter today (got it in an e-mail) and said this in about the MAX 8:

"Our experience with the MAX, along with the other U.S. operators, has been phenomenal. We've operated over 40,000 flights covering almost 90,000 hours. There is a ton of data collected, which we continuously monitor. In all of our analysis since our first flight in 2017, or that by our U.S. counterparts or the FAA; nothing has presented any flight safety concerns. It is also important to add that all Pilots at Southwest are deeply experienced and highly trained, as are our Mechanics who are highly experienced and trained to safely maintain every airplane in our fleet."

LarryJ

I am a pilot and have flown a variety of transport jets from the DC8 (designed in the 1950s), DC9 (1960s), B757/767 (1970/80s), B737 NG (1960-1990s), and now the B737 MAX (2010s).

Jets have a very wide operating envelope from low-speed, low-altitude to high-speed, high-altitude. Building an airplane that can operate throughout that wide range requires compromises. It can't be perfectly designed for any one phase of flight as that would make it impractical in other, still essential, phases of flight. All of the airplanes I've flown have had systems which provided "tuning/intervention" in certain phases of flight. The older designs did this mechanically. Newer designs rely more electronics and programming. The mechanical systems were far less reliable than today's electronic versions. None of them, unfortunately, can change the underlying physics which continue to make them an essential part of aircraft design.

Plato90s

Yes, I agree that modern fly-by-wire is better than the old mechanical systems.

By the same token, IT infrastructure has come a long way from the days of floppy disks, CD-ROM's, etc...

It doesn't change the basic idea of robust vs. fragile infrastructure. Those concepts existed even back in the days when mainframes ruled the computer world.

I already noted that even the robust infrastructure would end up requiring tuning/intervention. The key difference between robust and fragile is how often it had to be done, and the probability of a catastrophic cascade of failure if the tuning/intervention didn't work.



My understanding (as a non-pilot who isn't in aviation/aeronautic design) is that the 737-Max had a brand new software feature incorporated solely to manage the changes in how the aircraft flew. The behavior it was designed to counter (too-high angle of attack) is one which can lead to catastrophic failure.

In the IT world, if a vendor told me they added that kind of feature.... I'd consider it an indicator that there's going to be problems with robustness vs. fragility.



ETA:

In a way, this reminds me of the VW diesel emission scandal where after the problems were exposed, VW promised to fix it. Except they couldn't.

Subsequently, we learned that VW had been trying (secretly) for years to find a solution. Before the deception was exposed, VW went back to the urea system for diesel vehicle, an implicit admission that their "burn off" approach didn't work.

The engine was too fundamentally flawed and there was no way to change it to meet American emission standards. Eventually VW had to just recall and buy back the cars sold in the US to the tune of $8+ billion for ~400k vehicles.

Given that Boeing must have been working hard at trying to find a solution to the 737-Max problem since the Lion Air crash... are they going to end up like VW where it's just not viable to try to fix?

tmorse6570

It took two years to recover the recorder and they determined the probable cause to be a build up of ice crystals on a sensor. Kind of different.

LarryJ

As you said, we didn't know the cause for a couple of years. The airplane continued to fly with no resolution to the accident.

Productivity