LATAM 787 Sudden Drop due to 'Technical Incident'

[Evan]

Hmmm...

https://avherald.com/h?article=51601631&opt=0

A violent upset with 12 passenger injuries. This is a post taken from Aviation Herald (no idea if it is valid)

Power loss
By 787Capt on Monday, Mar 11th 2024 13:16Z

“In an airworthiness directive to be published Friday, the FAA said it is reacting to indications that “all three flight control modules on the 787 might simultaneously reset if continuously powered on for 22 days.”

“It said such a simultaneous reset in flight “could result in flight control surfaces not moving in response to flight crew inputs for a short time and consequent temporary loss of controllability".

That's not good.

There's a shade of the A330 upset incidents here. Those were caused by a hidden vulnerability to extremely rare ADIRU data spikes apparently the result of cosmic radiation, but the software was modified to prevent further occurrences. And I don't know if the 787 has a comparable pitch-lowering alpha protection function.

Whatever happened, it seems to be just as rare.

[flashcrash]

I don't know if the 787 has a comparable pitch-lowering alpha protection function.

Could this be a case of LATAM not following up on this AD relating to a fault in the B787 FCM? :

Federal Register :: Airworthiness Directives; The Boeing Company Airplanes

As I understand it, the AD (originally issued in 2016) requires 787 operators to reboot the FCM every couple of weeks because otherwise it'll eventually experience some kind of buffer overflow resulting in spuriously low airspeed data causing it to push the nose steeply down to correct a perceived stall? I have no doubt Boeing issued a fix, but perhaps LATAM hasn't installed it? The expected outcome does seem to closely match the unconfirmed pilot reports of what happened.

UPDATE: The article below however refers to a potentially more serious problem (as if this weren't bad enough already) in which power to the flight deck instruments is also lost if the aircraft's systems are not rebooted at least once every 240 days. I'm curious to know if the flight crew reported a sudden loss of instruments during this incident.

Power System Software Bug could bring down Boeing 787s - News (eepower.com)

[Evan]

AFAIK, there is no automated facility on the 787 with that kind of pitch (elevator, flaperons) authority other than the autopilot and the vertical gust suppression. It also gets airspeed from a synthetic airspeed algorithm, which, I suppose, could be effected by software vulnerabilities. The gust suppression is supposed to dampen turbulence upsets, not cause them, but perhaps a fault supplying data to this system is behind it. I'm expecting that, when we learn what caused this, it will be a deep dive into 787 system interdependencies. With very little technical documents available to the public, I'm looking forward to reading the report.

[Evan]

Could this be a case of LATAM not following up on this AD relating to a fault in the B787 FCM? :

Federal Register :: Airworthiness Directives; The Boeing Company Airplanes

As I understand it, the AD (originally issued in 2016) requires 787 operators to reboot the FCM every couple of weeks because otherwise it'll eventually experience some kind of buffer overflow resulting in spuriously low airspeed data causing it to push the nose steeply down to correct a perceived stall? I have no doubt Boeing issued a fix, but perhaps LATAM hasn't installed it? The expected outcome does seem to closely match the unconfirmed pilot reports of what happened.

UPDATE: The article below however refers to a potentially more serious problem (as if this weren't bad enough already) in which power to the flight deck instruments is also lost if the aircraft's systems are not rebooted at least once every 240 days. I'm curious to know if the flight crew reported a sudden loss of instruments during this incident.

Power System Software Bug could bring down Boeing 787s - News (eepower.com)

The GCU bug was revealed in 2015. Boeing was already removing the bug for a software update. The AD had 30-day compliance. Is it possible for a LATAM 787 to be powered up for 287 days, with no compliance on the AD and no software updates in almost nine years?! If it is, I'm going to have to put them on my list.

[flashcrash]

50 passengers injured, 12 hospitalized, 1 in "serious condition"

Fifty hurt as NZ-bound flight hit by 'technical' issue (bbc.com)

[TRTPUwU]

50 passengers injured, 12 hospitalized, 1 in "serious condition"

Fifty hurt as NZ-bound flight hit by 'technical' issue (bbc.com)

That is exactly why i always (even inflight on 13 hour flights) stay seated using my seatbelt. Some might say that is like missing the fun, but as a person who was injured by turbulence, i can assure it is worth it.

[BoeingBobby]

That is exactly why i always (even inflight on 13 hour flights) stay seated using my seatbelt. Some might say that is like missing the fun, but as a person who was injured by turbulence, i can assure it is worth it.

Let's see, I'm pretty damn sure that there is an announcement about keeping your belt fastened while seated. That the pilot "supposedly" told one of the passengers that he "lost control of the aircraft for a moment" is dubious at best.

[Evan]

Let's see, I'm pretty damn sure that there is an announcement about keeping your belt fastened while seated. That the pilot "supposedly" told one of the passengers that he "lost control of the aircraft for a moment" is dubious at best.

Well, he also told him it was his first day and promised that he would get the hang of it.

[BoeingBobby]

Well, he also told him it was his first day and promised that he would get the hang of it.

[Gabriel]

Let's see, I'm pretty damn sure that there is an announcement about keeping your belt fastened while seated. That the pilot "supposedly" told one of the passengers that he "lost control of the aircraft for a moment" is dubious at best.

I tend to believe him. Maybe not that he lost control, what I heard him say was the the pilot told him that he lost all instruments for a few seconds and then all came back. I can see how critical sensor information going "blank" for a few seconds can disable the instruments and the flight controls at the same time (in a fully FBW plane), since these sensor info are critical inputs for both the digital instruments and the flight control computers / FBW laws.

Not much evidence, but:
- The pax said he was surprised himself that the pilot told him that.
- Rumors that all 3 flight control computers self-rebooted at the same time.
- Rumors regarding existing procedures and AD's related to this, that include actions to reboot different systems every X hours, days or cycles, to prevent overflow errors.

Crazy stuff.

[BoeingBobby]

RUMOR is, A pilot never ever ever ever ever ever talks to ANYBODY. Until you have spoken to your Union representative and of course you are going to be talking to the chief pilot.

[Evan]

I tend to believe him. Maybe not that he lost control, what I heard him say was the the pilot told him that he lost all instruments for a few seconds and then all came back. I can see how critical sensor information going "blank" for a few seconds can disable the instruments and the flight controls at the same time (in a fully FBW plane), since these sensor info are critical inputs for both the digital instruments and the flight control computers / FBW laws.

Not much evidence, but:
- The pax said he was surprised himself that the pilot told him that.
- Rumors that all 3 flight control computers self-rebooted at the same time.
- Rumors regarding existing procedures and AD's related to this, that include actions to reboot different systems every X hours, days or cycles, to prevent overflow errors.

Crazy stuff.

But wasn’t that ‘bug’ also discovered many years ago? Hasn’t Boeing done anything to remove it? I mean, a vulnerability like that is not tolerable in anything you might want to call ‘airworthy’.

In the hands of a lesser crew… remember, a sudden loss of FCC’s caused (led to) the upset, stall and crash of an AirAsia A320.

[flashcrash]

But wasn’t that ‘bug’ also discovered many years ago? Hasn’t Boeing done anything to remove it?

I'm guessing those questions are rhetorical, but for the avoidance of doubt: Yes and yes. The info is in the text of the updated AD I provided a link to earlier. I know it's not an easy read, but pulling out the relevant parts:

A. The unsafe condition is described thus:

The FAA is issuing this AD to address deficiencies in the FCM software that could prevent continued safe flight and landing; to prevent unrealistic, sudden drops in displayed airspeed at high actual airspeed, which could lead to pilot control inputs that could exceed the structural capability of the airplane; to prevent simultaneous resets of all three FCMs, which could result in flight control surfaces not moving in response to flight crew inputs for a short time and consequent temporary loss of controllability; and to address potential unannunciated dual symmetric inboard slat skew, which can result in adverse handling characteristics of the aircraft.

B. There are three approved fixes:

(1) Use the onboard data load function (ODLF) to install FCM Block Point 3 software.
(2) Use the ODLF to install FCM Block Point 4 software.
(3) Install any later FAA-approved FCM software version using an approved method

C. The compliance date for applying one of those three fixes is:

Within 6 months after August 20, 2015

[Evan]

I'm guessing those questions are rhetorical, but for the avoidance of doubt: Yes and yes. The info is in the text of the updated AD I provided a link to earlier. I know it's not an easy read, but pulling out the relevant parts:

A. The unsafe condition is described thus:

The FAA is issuing this AD to address deficiencies in the FCM software that could prevent continued safe flight and landing; to prevent unrealistic, sudden drops in displayed airspeed at high actual airspeed, which could lead to pilot control inputs that could exceed the structural capability of the airplane; to prevent simultaneous resets of all three FCMs, which could result in flight control surfaces not moving in response to flight crew inputs for a short time and consequent temporary loss of controllability; and to address potential unannunciated dual symmetric inboard slat skew, which can result in adverse handling characteristics of the aircraft.

B. There are three approved fixes:

(1) Use the onboard data load function (ODLF) to install FCM Block Point 3 software.
(2) Use the ODLF to install FCM Block Point 4 software.
(3) Install any later FAA-approved FCM software version using an approved method

C. The compliance date for applying one of those three fixes is:

Within 6 months after August 20, 2015

So this doesn't add up. Unless LATAM hasn't done a ODLF update in almost nine years! That's unpossible...

But while we're here:

The FAA is issuing this AD to address deficiencies in the FCM software that could prevent continued safe flight and landing; to prevent unrealistic, sudden drops in displayed airspeed at high actual airspeed, which could lead to pilot control inputs that could exceed the structural capability of the airplane;

Sudden, unrealistic drops in airspeed are considered unreliable by competent pilots. The first order of business is to cross-check and if all three are in agreement, declare unreliable airspeed (most likely pitot icing), maintain current pitch and power settings, and work the QRH. A competent pilot will not respond to this by making 'pilot control inputs that could exceed the structural capability of the airplane' (or stall it). After AF447, hopefully this routine is well-rehearsed.

I would be more concerned about sudden unrealistic (yet still ruled valid) drops in airspeed effecting the avionics and the autoflight and/or gust suppression.

[3WE]

SUPPOSEDLY:

Hostie brings pilots “the fish.”

They bump a guarded switch on the back of the pilot seat.

The pilot seat begins moving forward.

Shoves the pilot into the control column.

Dive ensues.

Commentary:

It’s a completely different kind of MCAS, altogether.