Another Indonesian aircraft down

[Evan]

Which would be unfair, in this case. An engine can roll back for a number of reasons. FADEC, mechanical failures, bird ingestion, fuel contamination, the oil-fuel heat exchanger getting clogged, furl management issues, cnd autothrottle issues. Engine failures is probably the most-frequently practiced abnormal / emergency condition. We put 2 engines in the plane precisely in case one fails. If we cannot sustain flight with one engine, we would be better off with a single engine (since 2 engines double the probability of an engine failure).

Until we get to the fully autonomous planes, we need pilots, pilots that can manage some reasonable level of malfunctions, not paid passengers in the front seats.

Yes, unfair. But stealth, Gabriel, is the common factor I'm expecting. Not an abrupt, recognised asymmetry but a gradual one, building a gradual roll in IMC and, of course, pilot error, pilot disorientation, bias toward senses over instruments. But even if this turns out to be the case, the public can't understand these things. They will only understand that a computer malfunction 'caused' the crash.

[3WE]

Recurring Ignorant Outsider Asshat Safety Suggestions:

SOME sarcasm here, but a shred of seriousness.

Why not flashing yellow lights and a gentle hostie ring tone:

-Airspeed a little bit slow
-Engine power a bit unbalanced and wonky

We seem to be ass-uming that fat-dumb and happy flight suddenly went to hell (startle factor) and the stupid cowboy pilots botched it.

Several possible replies to this:

-There’s really nice and obvious airspeed and power indicators that the pilots might want to glance at.

-Similar (or adequate) systems already exist and we should STFU.

-In spite of incredible efforts, excrement transpires...we(no italics) can’t fix everything.

-Unintended consequences and 3BS’s gentle warning systems CAUSE distractions, confusion and crashes.

[BoeingBobby]

Despite having nothing in common with the -Max crashes, if the report finds that another computer-related failure led to pilot disorientation and upset, this really isn't going to play well for Boeing in the public opinion.

I knew it was only a matter of time. You ever hear of "click, click" fly the f*****g airplane?

[Evan]

I knew it was only a matter of time. You ever hear of "click, click" fly the f*****g airplane?

You ever hear of pilots neglecting to do that? You ever hear of pilots not noticing roll, pitch or altitude excursions when they assume the automation has it covered? You ever hear of complacency? You ever hear of distraction? You ever hear of disorientation? Yes, you have because these things are common causes of plane crashes.

I'm not blaming Boeing even if my speculations are true. I'm saying that the public probably will.

[Evan]

Why not flashing yellow lights and a gentle hostie ring tone:

There IS a flashing red light when there is an uncommanded disengagement of the autothrottle and the autothrottle is designed to disengage when the thrust levers (servo-driven, remember) become split. If it turns out that there was an autothrottle thrust asymmetry that did not involve split thrust levers or did not disengage because of them, that would be a dangerous malfunction. BoeingBobby rightfully points out that if such a situation arises, most crews will notice this and simply disconnect the automation.

But not all of them.

[TeeVee]

ok, bear with me here. i hear it's pretty common for pilots to take off and pretty much immediately engage HAL. let's assume they did that here, while also engaging the wonky auto-throttle.

wouldn't HAL have corrected for the asymmetrical thrust by adjusting control surfaces? if he did and it got to a point where HAL was designed to give up, wouldn't there have also been an alert saying HAL just turned tail and ran like the bitch that he is? would that have allowed the plane to "snap" into some uncontrollable attitude?

[Evan]

ok, bear with me here. i hear it's pretty common for pilots to take off and pretty much immediately engage HAL. let's assume they did that here, while also engaging the wonky auto-throttle.

wouldn't HAL have corrected for the asymmetrical thrust by adjusting control surfaces? if he did and it got to a point where HAL was designed to give up, wouldn't there have also been an alert saying HAL just turned tail and ran like the bitch that he is? would that have allowed the plane to "snap" into some uncontrollable attitude?

I think it's still fairly common for pilots to fly manually while navigating a departure but with the autothrottle engaged.

[3WE]

ok, bear with me here. i hear it's pretty common for pilots to take off and pretty much immediately engage HAL. let's assume they did that here, while also engaging the wonky auto-throttle.

wouldn't HAL have corrected for the asymmetrical thrust by adjusting control surfaces? if he did and it got to a point where HAL was designed to give up, wouldn't there have also been an alert saying HAL just turned tail and ran like the bitch that he is? would that have allowed the plane to "snap" into some uncontrollable attitude?

Nit picky: I wouldn’t say “snap” as it implies that someone paying attention would be helpless...


That being said, I very much concur, that Hal might have let go, they WEREN’T paying attention, and the plane “QUICKLY” went to hell...at least that’s a preliminary guess.

[Gabriel]

ok, bear with me here. i hear it's pretty common for pilots to take off and pretty much immediately engage HAL. let's assume they did that here, while also engaging the wonky auto-throttle.

wouldn't HAL have corrected for the asymmetrical thrust by adjusting control surfaces? if he did and it got to a point where HAL was designed to give up, wouldn't there have also been an alert saying HAL just turned tail and ran like the bitch that he is? would that have allowed the plane to "snap" into some uncontrollable attitude?

Yes, it can happen (and has happened in the past) that the AP tries to compensate some unbalance, and it does so for a while by adding control inputs until the point it gives up and at that point the controls get centered and stop compensating so you get for example a somehow suden roll. That has happened in the past and has caused accidents in the past and it may very well be part of what happened here.

If that happens, first you get a master warning aural (I think in the 737 is the "cavalry charge") and visual (master warning light in the dashboard right in front of each pilot) alarms when the autopilot disconnects. Then when you reach 35 degrees of bank (which is just 5 degrees more than normal, and very manageable) you will get an aural alarm saying "back angle, bank angle".

That of course doesn't guarantee that the pilot will either not react at all or, ore likely, will react in an irrational way like Air France or Colgan. In extreme cases it can also happen that the pilot reacts rationally but inadequately due to, well, the case being extreme, and the lack of training in that specific failure (the rudder hardover cases of United at Colorado Springs and USAir at Pittsburg come to mind).

Yet in other cases, the pilots put more focus in trying to reconnect the automation than in trying to fly the plane themselves (which is the correct thing to do when the automation is failing).

[TeeVee]

i'm gonna go out on a limb here and say it was more than simply a wonky auto-throttle. clearly all twins have the ability to counter asymmetrical thrust up to the setting to maintain flight at takeoff and then some. the autopilot can deal with huge crosswinds at cruise, as in flying north south in the US in the jetstream.

so unless the auto-throttle engaged reverse thrust...

[Gabriel]

i'm gonna go out on a limb here and say it was more than simply a wonky auto-throttle. clearly all twins have the ability to counter asymmetrical thrust up to the setting to maintain flight at takeoff and then some. the autopilot can deal with huge crosswinds at cruise, as in flying north south in the US in the jetstream.

so unless the auto-throttle engaged reverse thrust...

Not sure what parts of your post are serious and which ones are sarcastic, so I'll respond to it as if it was all serious.

- Crosswind at cruise, no matter how huge, has absolutely zero to do with either of asymmetric thrust or rudders.
- Autothrottles have no way whatsoever to engage reverse thrust in a 737. Due to how the reverse thrust system is designed, no failure mode of the autothrottles has a mean to engage them.
- Twins have the ability to deal with full thrust on one engine and zero thrust in the other, and pilots have the ability to screw it up.

I don't know what brought down this plane, but asymmetric thrust due to autothrottle malfunction triggering a totally inapropiable response by the crew is not out of the realm of possible scenarios that fit the little that is known so far. It should that happen, but then a pilot should also not relentlessly pull up and actively stall the plane as the first immediate response to a stickshaker (Colgan) or, upon an unreliable speed event at cruise, respond by pulling up 1.5Gs into a 7000 feet per minute climb climbing 2500 ft and, when the plane finally stalls at the top of the crazy climb and the stall warning starts shouting "stall stall" pull up again, fully stall the plane, and keep pulling up all the way into the Atlantic Ocean.

Of course, the above doesn't make what you say impossible or crazy (that there was more than a wonky autothrottle coupled with horrible pilot performance)

[TeeVee]

only sarcasm was going out on the limb.

i posted that after speaking to my friend (ex-737 pilot, though the 800), who confirmed that a wonky auto-throttle could not "over-power" flight controls. if it could, that would mean that single engine flight would be essentially impossible. the mention of cruise flight was simply to point out that the flight controls have more than enough effect to counter even very high velocity crosswinds found at cruise altitudes.

i guess what i'm getting at is this: though what evan says about public opinion is likely correct, i dont think this is gonna be a design flaw/failure case that boeing will be to blame for.

[3WE]

Asymmetric thrust CAN AND TOO OFTEN DOES (see footnote) over power controls...you simply have to be going very slow...(so, imagine climbing and one engine slowly powers back and the speed reduces).

Footnote: Too often in GA, rare in airliners due to intense recurrent training, and more robust designs.

”There must be more to this”. Indeed, there usually are additional factors, and to simply say “auto throttles” is wrong.

Sadly, there are ongoing cases of folks crashing planes over minor problems...(there’s a reasonable argument that MCAS is a minor problem- This loud huge ass trim wheel is acting strange, let’s turn it off.)

Throw in the human factors game...in spite of extremely robust designs, you burn out a light bulb and crash into the Everglades, or use fundamental HOW TO STALL PROCEDURES to combat a several-minute, 36,000 ft descent...

[Evan]

Asymmetric thrust CAN AND TOO OFTEN DOES (see footnote) over power controls...you simply have to be going very slow...(so, imagine climbing and one engine slowly powers back and the speed reduces).

Followed by stickshaker, followed by pressing the toga button, followed by asymmetic toga thrust, not followed by manual rudder inputs (no yaw channel on 737-classic autopilots) or appropriate pilot responses...

[Gabriel]

(no yaw channel on 737-classic autopilots)

But they do have yaw damper, don't they?

I do wonder if the AP / yaw-damper are capable of handling a high thrust-asymmetry situation without pilot input on the pedal or rudder trim.