General > General Technical Chat

And now the Boeing 777...

<< < (10/15) > >>

Monkeh:

--- Quote from: ve7xen on March 01, 2021, 04:45:48 am ---
--- Quote from: Monkeh on February 27, 2021, 08:14:16 pm ---
--- Quote from: drussell on February 27, 2021, 07:56:43 pm ---Airbus = Awesome

--- End quote ---

AF447.

--- End quote ---

 |O If ever there were an accident where it was appropriate to say 'it was completely pilot error', this is it. There's always blame to go around, and there were some HMI problems identified with the Airbus flight controls, but you really can't blame the airframe or control systems for a pilot continually doing the opposite of what they were trained to do and the procedures say to do for multiple minutes while they fly the plane into the sea. The pilots still have mostly ultimate authority, and if what they are actively doing is going to crash the plane, with the exception of alpha-prot, the airframe isn't going to stop them.

--- End quote ---

I attribute a degree of blame to Airbus because the pilots (neither particularly good, clearly) were unable to accept they were in a stall because it's an Airbus, and it won't let you stall. Except when it does.

And yes, it was pilot error. Like most pilot errors, there are contributing factors.

ve7xen:

--- Quote from: Monkeh on March 01, 2021, 04:52:23 am ---I attribute a degree of blame to Airbus because the pilots (neither particularly good, clearly) were unable to accept they were in a stall because it's an Airbus, and it won't let you stall. Except when it does.

And yes, it was pilot error. Like most pilot errors, there are contributing factors.

--- End quote ---

Usually I'm with you on there being multiple critical factors to blame, I just don't really think that was the case in this accident to the degree it usually is. They identified the unreliable airspeed early on, and understood it later in the sequence, yet never completed the associated procedure nor were their control responses appropriate for that condition. They didn't discuss the stall alarms at all, which is pretty strange, but doesn't lead to the conclusion that they thought they could ignore the continuous stall alarm because they were in normal law; they also called out the shift to alternate law, though I agree it could have easily been missed. There's no excuse for not executing the stall recovery procedure (or even earlier, the unreliable airspeed procedure - some of the items were performed but it was not done in the expected challenge/response systematic manner, nor completed), nor for the PNF to continually provide control inputs after giving the controls to the Captain - or this was a CRM failure where they had conflicting interpretations of who was in control, but regardless it is a piloting issue.

Confusion in complex failures is a relatively common cause of accidents, and I'm usually the first to defend the pilots in a situation with conflicting alarms and information, but in this case the only alarm they were getting initially was STALL STALL, and their conditioned response to that should be to execute a stall recovery or if they were actually looking at their PFD, maybe unusual attitude recovery could be justified. And the only unreliable indication they were receiving was airspeed, which recovered shortly after the incident began. It appears they just completely ignored the attitude indicator they should have known they could trust, as well as the repeated stall warnings. Suspecting unreliable airspeed seems to have led them to mistrust the airplane entirely and ignore everything else it was telling them, and that's not defensible, especially in hard IMC.

The only places where the aircraft could arguably have done a better job IMO would be not suppressing the stall alarm below 60KIAS, and kicking the flight directors completely off when the data was invalid, rather than restoring them as the data came back, but by the time the incident evolved that far, they should have already executed the stall recovery procedure. Even after it did get that far, getting out of stall and back to a normal attitude should be overriding anything else in their minds. Regardless of anything else, including airspeed which was suspect, it would have been obvious they were at an extreme nose up attitude that would lead to stall. You could argue the pilots should have visibility of AoA data, but this isn't unique to Airbus, and they certainly didn't need it to tell them they were stalling the plane, the plane did that all on its own, as would have their attitude indicator.

The BEA didn't request any changes to the airplane as a result of the accident, and I think that's pretty clear indication they don't think any part of it was causal here. It's just bad pilot decision making from the Captain's laissez faire attitude about the weather, to not executing the unreliable airspeed procedure, to ignoring the stall warnings, to applying inappropriate control inputs, to continuing to apply those inputs after the Captain ostensibly took control.

Anyway, apologies for dragging this off topic, just get tired of people bringing up AF447 as if it's clearly the fault of Airbus; even with the most generous interpretation, the airplane plays a minor part in this accident.

Cubdriver:

--- Quote from: ve7xen on March 01, 2021, 08:42:13 pm ---
--- Quote from: Monkeh on March 01, 2021, 04:52:23 am ---I attribute a degree of blame to Airbus because the pilots (neither particularly good, clearly) were unable to accept they were in a stall because it's an Airbus, and it won't let you stall. Except when it does.

And yes, it was pilot error. Like most pilot errors, there are contributing factors.

--- End quote ---

Usually I'm with you on there being multiple critical factors to blame, I just don't really think that was the case in this accident to the degree it usually is. They identified the unreliable airspeed early on, and understood it later in the sequence, yet never completed the associated procedure nor were their control responses appropriate for that condition. They didn't discuss the stall alarms at all, which is pretty strange, but doesn't lead to the conclusion that they thought they could ignore the continuous stall alarm because they were in normal law; they also called out the shift to alternate law, though I agree it could have easily been missed. There's no excuse for not executing the stall recovery procedure (or even earlier, the unreliable airspeed procedure - some of the items were performed but it was not done in the expected challenge/response systematic manner, nor completed), nor for the PNF to continually provide control inputs after giving the controls to the Captain - or this was a CRM failure where they had conflicting interpretations of who was in control, but regardless it is a piloting issue.

Confusion in complex failures is a relatively common cause of accidents, and I'm usually the first to defend the pilots in a situation with conflicting alarms and information, but in this case the only alarm they were getting initially was STALL STALL, and their conditioned response to that should be to execute a stall recovery or if they were actually looking at their PFD, maybe unusual attitude recovery could be justified. And the only unreliable indication they were receiving was airspeed, which recovered shortly after the incident began. It appears they just completely ignored the attitude indicator they should have known they could trust, as well as the repeated stall warnings. Suspecting unreliable airspeed seems to have led them to mistrust the airplane entirely and ignore everything else it was telling them, and that's not defensible, especially in hard IMC.

The only places where the aircraft could arguably have done a better job IMO would be not suppressing the stall alarm below 60KIAS, and kicking the flight directors completely off when the data was invalid, rather than restoring them as the data came back, but by the time the incident evolved that far, they should have already executed the stall recovery procedure. Even after it did get that far, getting out of stall and back to a normal attitude should be overriding anything else in their minds. Regardless of anything else, including airspeed which was suspect, it would have been obvious they were at an extreme nose up attitude that would lead to stall. You could argue the pilots should have visibility of AoA data, but this isn't unique to Airbus, and they certainly didn't need it to tell them they were stalling the plane, the plane did that all on its own, as would have their attitude indicator.

The BEA didn't request any changes to the airplane as a result of the accident, and I think that's pretty clear indication they don't think any part of it was causal here. It's just bad pilot decision making from the Captain's laissez faire attitude about the weather, to not executing the unreliable airspeed procedure, to ignoring the stall warnings, to applying inappropriate control inputs, to continuing to apply those inputs after the Captain ostensibly took control.

Anyway, apologies for dragging this off topic, just get tired of people bringing up AF447 as if it's clearly the fault of Airbus; even with the most generous interpretation, the airplane plays a minor part in this accident.

--- End quote ---

The problem I have with the Airbus design with its decoupled side sticks is perfectly illustrated in this incident - the pilot flying is mistaken, and is inputting full nose up on his control stick.  The guy on the other side of the cockpit doesn't realize this, because he can push his stick full forward independently.  The aural warning says something about the mixed control inputs, but that can be missed in the confusion.  Boeing's control columns are coupled together - the guy in the right seat is going to know what the guy in the left seat is doing, because his control wheel moves in unison.  From what I've read of that accident, it seems that at times the left seater was trying to lower the nose, but the right seater had control authority and so it stayed up.  There'd have been no question how the controls were positioned in a Boeing craft, as BOTH the yokes would have been pulled back, and you can FEEL if the guy next to you is fighting you on the controls.  Seems to be a fundamental design flaw in the controls that can rise up under confusing/stressful situations to bite you hard in the ass, permitting erroneous control inputs where one side has no idea what the other is doing, as we saw here.

-Pat

<edit - switched seats>

Monkeh:

--- Quote from: ve7xen on March 01, 2021, 08:42:13 pm ---Anyway, apologies for dragging this off topic, just get tired of people bringing up AF447 as if it's clearly the fault of Airbus; even with the most generous interpretation, the airplane plays a minor part in this accident.

--- End quote ---

It's not clearly the fault of Airbus, I just don't hold that "Airbus = awesome". The whole concept of preventing the pilot from making improper control inputs and applying flying skill right up until the moment those skills are necessary is, in my opinion, a mistake, and one I believe contributed to that incident, among others. AF296, perhaps - yes, pilot error, compounded by "the computer will protect us" (the whole point of the demonstration) and "oh crap, the computer won't let me pull up".

james_s:

--- Quote from: Cubdriver on March 01, 2021, 09:56:30 pm ---The problem I have with the Airbus design with its decoupled side sticks is perfectly illustrated in this incident - the pilot flying is mistaken, and is inputting full nose up on his control stick.  The guy on the other side of the cockpit doesn't realize this, because he can push his stick full forward independently.  The aural warning says something about the mixed control inputs, but that can be missed in the confusion.  Boeing's control columns are coupled together - the guy in the right seat is going to know what the guy in the left seat is doing, because his control wheel moves in unison.  From what I've read of that accident, it seems that at times the left seater was trying to lower the nose, but the right seater had control authority and so it stayed up.  There'd have been no question how the controls were positioned in a Boeing craft, as BOTH the yokes would have been pulled back, and you can FEEL if the guy next to you is fighting you on the controls.  Seems to be a fundamental design flaw in the controls that can rise up under confusing/stressful situations to bite you hard in the ass, permitting erroneous control inputs where one side has no idea what the other is doing, as we saw here.

--- End quote ---

I do very much prefer the Boeing cockpit layout. Although I am not a pilot myself, mechanically coupling the controls on both seats is one of those things that just seems like an obviously good idea. I also think a yoke is a much more logical input device for something like an airliner than a stick, but again I'm not a pilot so I have not experienced either one outside of PC simulators.

Navigation

[0] Message Index

[#] Next page

[*] Previous page

There was an error while thanking
Thanking...
Go to full version
Powered by SMFPacks Advanced Attachments Uploader Mod