[rob74]

Even with zero controls, and even if one or more of the various computers decided spontaneously to restart, I would expect the plane to continue flying the way it did before the incident rather than going into an (apparently) uncommanded descent? I mean, we had that with the 737 MAX, so I wouldn't rule it out, but it sounds suspiciously like the pilot messed up and is trying to blame the airplane. However I'm no specialist, so it's probably best to wait until further details emerge...

[thombat]

But what control input could a pilot make that abruptly produces significant negative gees for just a couple of seconds? Other than the bland references to "technical issue" it sounds like clear air turbulence. (Although one possible mistake could be that the weather radar did warn of it and the pilot didn't react?)

[MadnessASAP]

An autopilot disconnect (caused by a failure of the flight control computer) in an out of trim condition can cause it. I'm not sure how a 787s trim system is built but on our, much older, plane with a mechanical trim system there's a motor that will slowly move the trim tabs to bring the autopilot inputs to 0.

That is if the autopilot is producing a constant nose up control signal the auto-trim will move the elevator tabs towards nose up until the AP pitch signal is null.

It's to prevent fun excursions like this should the autopilot become disconnected without the pilots hands on the controls.

[histriosum]

Weather radar cannot see clear air turbulence. Essentially you are relying on forecasts and pilot reports.

[hugh-avherald]

From reports of passengers in the rear of the plane suffering the most injuries, it sounds like a descent followed by an abrupt pitch up (elevators push tail down).

[kayodelycaon]

Airbus had a problem like this on an A330 due to corrupted AOA data in the flight computers.

https://en.m.wikipedia.org/wiki/Qantas_Flight_72

[rvba]

Thanks for the link, it made me move to the article about "Mayday Mayday Mayday" callsign, then listen to a recording of one and read about a ship that hit a bridge in Florida. They rebuild the bridge.

Wikipedia is such a fascinating website

[sammy2255]

Cosmic rays*

[kayodelycaon]

The cause of the corrupted data is not known but it did happen multiple times in a short period and only on an A330s.

Once Airbus knew what the problem was, they were able to detect and mitigate it with a software update. They didn’t fix the hardware.

[throwaway3563]

The most that the ATSB was able to determine was that the data corruption was basically akin to a C++ reinterpret_cast of "altitude" as "angle of attack", causing the 37,000ft or so altitude to sporadically be read as a 50 degree AoA.

The issue was not definitively traced to cosmic rays or another root cause.

[cm2187]

Well, something of that sort happened with the A330 also: https://en.wikipedia.org/wiki/Qantas_Flight_72