If this is a bit flip, shouldn’t these devices be tested before service??
Yes, it can be expensive otherwise: http://www-users.math.umn.edu/~arnold/disasters/ariane.html
Peter wrote:
You would think a 22 degree AoA error would become apparent elsewhere if it was a software thing e.g. corruption of a memory location.
The single sensor failure is probably not a big problem if MCAS was limited to say +/-2 degrees trim/pitch on that 22 degrees AoA reading error in some “emergency mode”?
Another viewpoint here.
And here.
Of course, a European is likely to say that these are the US pilots patriotically defending a US company. But if you read the specifics, it’s interesting, and there are valid points. Autoflight systems can fail in various ways, including catastrophically, and pilot must know how to deal with these (so e.g. an Airbus has the reversionary modes, including all the way back to losing all the flight computers). The Q boils down to whether a pilot should have been aware of how to disable the system. These articles claim that the answer is Yes, and assert that “3rd World” pilots involved in the two crashes apparently didn’t.
Boeing must be exceedingly happy with this statement.
It’s not that an undocumented, unprecedented automatic trim system, which doesn’t move continuously, which at first glimpse can be manually overridden (only to clandestinely increase the mis-trim after 5 sec) is the factor here, no, plain simply the sauvages don’t know how to fly.
This interview with quite well known aircraft designer Jan Roskam touches the topic briefly on page 34.
One big issue here is that MCAS is active only after the crew has cosciously switched off all automation to hand fly. And it is NOT primarily a stall prevention system (although often characterised as such), like a stick pusher where activation is obvious, but an augmentation system that aims at increasing the control force required in a subtle way so the aircraft handles like the NG.
So while perhaps a better crew could have recognised the trim runaway and switched the thing off – the one on the flight just prior to the Lion Air crash did – this feature is stealthy, and the difference in safety record so stark that better training alone is not the answer
Peter wrote:
Of course, a European is likely to say that these are the US pilots patriotically defending a US company
Not that we have not seen the same in Europe….
Peter wrote:
Airbus has the reversionary modes, including all the way back to losing all the flight computers
It does indeed. However, none of the protections an Airbus has will ever fly the airplane into the ground. Alpha floor by Airbus is a very different protection than this thing. The closest to what Boeing came up with here (or maybe where they got the idea from) was a similar system in the MD11, but that one obviously worked better and was well documented. I am not aware of a single event where it caused a massive upset.
Cobalt wrote:
And it is NOT primarily a stall prevention system (although often characterised as such), like a stick pusher where activation is obvious, but an augmentation system that aims at increasing the control force required in a subtle way so the aircraft handles like the NG.
Yes but with one significant difference:
If there is trim movement of any sort in an NG or before and the control yoke is moved in the opposite direction, the trim movement stops immediately. So nose down trim, pull, and it stops.
Not in the Max. That feature called trim brake is gone. So if it starts to trim nose down and you pull, it will continue trimming down until you run out of elevator authority.
Mooney_Driver wrote:
The closest to what Boeing came up with here (or maybe where they got the idea from) was a similar system in the MD11, but that one obviously worked better and was well documented. I am not aware of a single event where it caused a massive upset.
Yes that is how we got an upgrade DC10 to MD11, but the difference DC10/MD11 was well documented and understood before the aircraft was in the sales book, so the few accidents from MD11 were mainly on gusty cross-wind landings rather than falling from the sky after passing FL100 with flaps up…
Hand flying something inherently unstable/by-wire when the computer is off requires at least some documentation/training?
Also, even with good understanding of the system, I am not sure why we expect a fresh/experienced airline pilot can hand fly a Typhoon? (that is probably that is what you will get when you disable MCAS on B737Max? assuming you manage to locate the right one cutout switch out of 3 on time…)
http://www.airlinesafety.com/faq/faq9.htm
Last section has usual anti-Airbus BS: “Somebody once said they should let Lockheed design all the airplanes, Boeing build them…and McDonnell-Douglas market them! And let the French guys stick to making Citroens and Peugeots…”, so one will need to take it with a pinch of fromage 
Ibra wrote:
Also, even with good understanding of the system, I am not sure why we expect a fresh/experienced airline pilot can hand fly a Typhoon? (that is probably that is what you will get when you disable MCAS on B737Max? assuming you manage to locate the right one cutout switch out of 3 on time…)
That is just a silly comment. It has nowhere been stated the aircraft can’t be handflown. They are trying replicate the NG handling. That uses an auto-trim system which is clearly poorly designed. This led to the problems when a sensor failed.
