The Other Recent Deadly Boeing Crash No One Is Talking About (nymag.com)

New York magazine's Intelligencer remembers last month's crash of a Boeing 767 carrying cargo for Amazon and the U.S. Postal Service -- and shares a new theory that its cause wasn't a suicidal pilot or an autopilot malfunction: In online pilot discussion forums, a third idea has been gaining adherents: that the pilots succumbed to a phenomenon called somatogravic illusion, in which lateral acceleration due to engine thrust creates the sensation that one is tipping backward in one's seat. The effect is particularly strong when a plane is lightly loaded, as it would be at the end of a long flight when the fuel tanks are mostly empty, and in conditions of poor visibility, as Atlas Air 3591 was as it worked its way through bands of bad weather. The idea is that perhaps one of the pilots accidentally or in response to wind shear set the engines to full power, and then believed that the plane had become dangerously nose-high and so pushed forward on the controls. This would cause a low-g sensation that might have been so disorienting that by the time the plane came barreling out of the bottom of the clouds there wasn't enough time to pull out of the dive.

It has been speculated that this might have been the cause of another bizarre and officially unsolved accident from three years ago: Flydubai Flight 981, which crashed 2016 in Rostov-on-Don, Russia.... While it's still too early to draw any kind of conclusions about Atlas Air 3591, the possibility exists that a firm conclusion will never be drawn -- and if it is, the cause could turn out not to be a design flaw or software malfunction that can be rectified, but a basic shortcoming in human perception and psychology that cannot be fixed as long as humans are entrusted with the control of airplanes.

Comments

Flight control and it's not a design error

[Anonymous Coward]

First, I'm not trying to justify Boeing or what they did. Every indication is that they f**ked up royally.

However, the reporting on this by the non-technical media has gotten out of control and is nothing short of "Boeing built and sold a deathtrap". They did not and it was not just Boeing's failures (though Boeing could have done a lot more to eliminate the failures by others... or even the possibility of them).

The 737 MAX MCAS (Maneuvering Characteristics Augmentation System) improves the flight handling of the 737 MAX which has different flight characteristics from other 737's because the engines (which are by far the heaviest component on the plane) were moved ~6 inches forward of the wing to accommodate the larger engine size. This has often been characterized in the media as meaning the plane is unstable or not flight worthy (both incorrect) and that the MCAS is a "hack" that was snuck past the FAA (also incorrect). MCAS systems are very common and have been for decades. There are planes that are unstable in flight and CANT be flown without an MCAS but these are generally military aircraft where the increased agility or other benefits justify the risk (the F-117 is a perfect example of this, the shape required for that generation of "stealth" created an aircraft that was unstable).

What seems to have lead to both the Lion Air and Ethiopian Air crashes was not the MCAS per se but the way the plane responds to a stall event perceived by the angle of attack sensors (AOA). On all 737's ("original", "NG" and "MAX") the MCAS will put the nose of the plane down in attempt to prevent the stall. On every generation before the MAX this behavior could be canceled by the pilot by pulling back on the yoke (which is intuitive... the pilot is directly counter-acting the nose down). On the MAX this behavior was changed, the MCAS had to be turned off by pulling a circuit breaker. This was Boeing's first screw up, that is not intuitive and a break from decades of previous behavior. (Note: this type of circuit breakers are not like the ones in your car where you have to go digging for them, they are switches that are readily available to the pilot).

Where Boeing further screwed up (and can/may be viewed as criminal subterfuge) is that they did not make this information expressly clear to _everyone_. From everything I've read it was clearly documented in the maintenance manuals for all airlines but it was not clearly documented in the pilot's manual. For some airlines it was, for some it wasn't. The question is wether or not Boeing did this so that the plane would qualify as the same type as the "NG" meaning pilots qualified for the NG would be qualified for the MAX (which saves training/SIM costs for the airlines and also makes it easier for them to schedule/rotate pilots).

IF both crashes were caused by invalid responses (nose down) to a faulty AOA sensor then both crashes were 100% preventable by the pilots. This is clearly demonstrated by the fact that the Lion Air aircraft had the same incident occur the previous day and a crash was averted because a dead-heading pilot in the cockpit jump-seat did know about the change in behavior and advised the pilots to pull the circuit breaker. The flight continued without incident.

So no, the MAX does not have major design flaws. When handled correctly, it flies like it's supposed to. It's notable that Ethiopian Air (and all other airlines it seems) immediately updated their 737MAX training after the Lion Air crash brought the difference to light. The pilot in command of the EA flight that crashed skipped this updated training (I have not seen any information about the co-pilot).

The root issue is not the design changes on the MAX, it's how the design changes were communicated:
1. Boeing should have made the new requirement to disable the stall handling much more clear.
2. The FAA probably should not have approved it as the same "type" as a 737NG.
3. The Airlines should have been more diligent in their assessment of th