AF447 Vanity Fair article

Jimflyfast said:
I agree with most of what you are saying here

The problem was pairing a weak pilot(s) with an airplane that did something some engineers swore it could never do. When you tell someone a ship cant sink, then the captain doesnt slow down near icebergs. If you tell a pilot that a plane can not stall, it is not in his mind that is a possible outcome. It isnt even in the back of his mind a something he should be recovering from. The airbus design of invalid airspeed below 100 knots in cruise was a fatal assumption. What would you do if when you pulled back, the plane was quiet, and when you pushed, it suddenly said "Stall"? Throw in night, likely no horizon, startle and uncertainty, and damn near everyone would do the same thing. The autotrim was the part that sealed their fate. The nose was trimmed full up the entire descent. Even if he had pushed, it was unlikely he would have held it long enough for the trim to run the correct way. I would say that passing 20000 feet they were dead, as the only recovery from that high of a deck angle was full rudder to drop the nose, and there are about 4 line pilots in the world that are willing to use full rudder in cruise.

I could go on and on about this, either as an engineer and design assumptions or as someone with a fair enough time in the 330, or as someone who has done full aerodynamic stalls in transports at altitude
Click to expand...

Woah. I don't even know where to start after just reading this. The more that I learn about Airbus design philosophy, the more it honestly scares me more and more, but this whole autotrim thing is perhaps in another league?

As someone who's enjoyed an AoA instrument (Metroliners, Lears, Falcons- and yes, I'm aware that the former 2 don't read actual angle), I find them incredibly useful, and wish that all aircraft would have a real, calibrated AoA instrument.
 
Patrick said:
Woah. I don't even know where to start after just reading this. The more that I learn about Airbus design philosophy, the more it honestly scares me more and more, but this whole autotrim thing is perhaps in another league?

As someone who's enjoyed an AoA instrument (Metroliners, Lears, Falcons- and yes, I'm aware that the former 2 don't read actual angle), I find them incredibly useful, and wish that all aircraft would have a real, calibrated AoA instrument.
Click to expand...

It was auto trimmed full nose up because that's what the PF was commanding it to do. The trim was not a malfunction.
 
Patrick said:
Woah. I don't even know where to start after just reading this. The more that I learn about Airbus design philosophy, the more it honestly scares me more and more, but this whole autotrim thing is perhaps in another league?
Click to expand...

There's nothing really strange about auto trim. Basically the flight controls trim into position so you don't have to hold the stick back to keep the nose up. It's how every flight simulator (computer game) ever made works. Holding the stick back would command the trim to keep running (much like holding a trim switch down) until it hit the stop.
 
MikeD said:
While 3000 hours isn't low total time at all. With his focused, ab initio background, he potentially could run into a situation where he, in reality, has 1 hour, 3000 times. I don't know what kind of flying he did apart from only Air France, but there could variations on that experience/quality of time theme being somewhere in between the two extremes?
Click to expand...

My thinking as well. Tempting for an ab initio guy - or gal - to do the bare minimum in basic flight training. "Why focus on stalls, etc.? I'm never going to use them in my shiny new Airbus". The question isn't about deterioration of skills from automation - it's whether they had such skills in the first place.

The captain's background was fine, but he was off the premises until it was almost too late, and operating on one hour of sleep.

Langewiesche did some "psych" type analysis of cockpit interactions here, and did so with the Lear/Boeing Brazil accident as well. Interesting, but I couldn't comment on the accuracy of same without hearing the actual recordings - transcripts aren't enough.
 
From both of these articles (I read the one from @seagull link), I gained a much better understanding of this flight control system than I had when I commented on page one of this thread.
Knowing what I know now, I must state, "TERRIBLE flight control design too......IMHO...:confused2:"
 
MikeD said:
AOA gives you instant visual of best range/endurance for cruise, as well being used on approach/landing for being on appropriate speed at any given time. Far more accurate than airspeed indications for certain flight conditions, and used almost exclusively during landing in tactical jets.
Click to expand...

The vast majority of "speeds" in pretty much every realm of flight in the F-15E (to include dogfighting) reference AOA and not an airspeed. Airspeeds can be manually computed, but the AOA is the true technical order reference.
 
seagull said:
My point is that the F/O with 3000 hours is not all that "low time". Also, I am fairly sure his stick and rudder skills were just fine, as were the AA 903 and the C-5 crew I discuss in my article. As for going through the articles, those are just off the top of my head, there were other problems in it.
Click to expand...

Having spent time flying with a good variety of C-5 (and other USAF heavy) pilots when we were both transplanted into a third aircraft were were somewhat new to (King Air), I would argue that their stick and rudder skills were probably not "just fine" if they were typical. Most USAF heavy guys suffer from the same automation-induced atrophy of basic stick-and-rudder skills that 121 airline pilots can. The USAF mandates and standardizes the very same use/reliance on FMS and FGS/autoflight systems in the heavy world with the same rationale of safety and efficiency, and there is a predictable result on pilot basic hand flying skills and SA.

I was an instructor in the King Air, and guys would get into theater with lots of hours in their designated MWSs, but only a dozen or so hours in the King Air. When I would take them on their initial sorties in theater, I would mandate that they hand fly pretty much everything on the sortie outside of the actual mission orbit. You would be surprised how many pilots were extremely apprehensive about doing this, and they subsequently made the same types of execution errors that you see in basic flying training students; overcontrolling, undercontrolling, chasing airspeeds and altitudes and unable to maintain stable flight, failure to trim, lack of SA on all kinds of things. It was ugly....and a little humorous...and even a little sad. Most of them were pretty embarrassed and even got angry with me for "setting them up" (or any number of rationalizations for their pretty crappy hand flying). They would have a rough flight or two, and then they'd have the rust knocked off and be back to some semblance of actual flying skills.

This wasn't just my experience in this single situation, either. When I was an instructor at a training command base, there were a whole ration of heavy MWS pilots (the ones I'm thinking of were specifically C-17, EC-130, and KC-135 pilots) who had significant difficulty passing the instructor qualification for the basic trainer, the T-6. Yet again, their hands weren't as well "tuned" and their brains weren't working at a fast enough pace to make single-seat airmanship decisions (not to mention the types of decisions they'd have to make at the next level as actual instructors). This wasn't the norm for heavy MWS-sourced instructors, but just for comparison there were no fighter or bomber background pilots who were having these kinds of stumbles learning how to instruct in a basic trainer.

None of this is because these pilots were idiots. These were all highly experienced, very knowledgeable, highly trained pilots who had made it through years of extremely competitive and selective programs to just become a military pilot in the first place. It was not a "talent" problem, but an "experience" problem. Stick-and-rudder is a perishable skill, and if you spend thousands of hours in the Flight Levels with the autopilot on, or flying STARs and SIDs using the FGS/autoflight, the connection between the brain and the hands to make the yoke and thrust levers do what you want them to do isn't as sharp as it may have been in the past (or, sadly, ever even been!). The connection between the eyes and ears and the decisionmaking part of the brain atrophies just as wickedly.

So, in the grand scheme, we are talking about a pilot's ability to handle a nonstandard situation using basic airmanship. As noted above, guys with automation-induced atrophy eventually got their groove back after some knocking off of the rust...but when you are talking about handling an emergency, there really isn't time for that recurrency to take place. It has to be sharp at the very instant the pilot needs to call on those skills to handle the emergency.

To put this in perspective, though, I suppose you have to qualify what it means to say that these pilots' "stick and rudder skills were just fine". Their skills may have been "just fine" enough to handle normal operations in the transport category utilizing primarily FMS and FGS systems to operate the aircraft. To me, that is not the correct measurement.
 
Patrick said:
but this whole autotrim thing is perhaps in another league?
Click to expand...

Nothing particularly weird or insidious about autotrim. It simply has some limitations and impacts on different flight regimes that have to be kept in mind and managed/counteracted.
 
Hacker15e said:
Nothing particularly weird or insidious about autotrim. It simply has some limitations and impacts on different flight regimes that have to be kept in mind and managed/counteracted.
Click to expand...
For the record, I wasnt implying that the autotrim was bad or that it failed. It worked as designed, the pilot was calling for slower speed by pulling up continuously. It just completely sealed their fate.

Autotrim is the only way a sidestick control FBW could possibly work, as there is no stick force to indicate out of trim.
 
jynxyjoe said:
Don't even remember that on the CRJ (except for a autopilot trim and the like which isn't the same).
Click to expand...

It autotrims for Mach, flaps, and autopilot if I recall. I thought auto-trim was auto-trim? They just don't schedule it to function during normal hand flying...?
 
cencal83406 said:
It autotrims for Mach, flaps, and autopilot if I recall. I thought auto-trim was auto-trim? They just don't schedule it to function during normal hand flying...?
Click to expand...
Well if you're handflying and not trimming the plane will just nudge it for you after a few seconds. It's like having a little buddy along making sure you're not completely and stupidly out of trim at any time.

I think the CRJ leaked out of my ear quickly... but I still can draw the elect system for the Beech 1900, so I have that... which is nice.
 
jynxyjoe said:
737. Don't trim as you accelerate, fun to watch it fix it for you. :)
Click to expand...
The 737 Speed Trim System is not similar to autotrim in the airbus, only works in specific modes, can be overridden, and it is obvious that it is happening by seeing and hearing the trim wheels move. In most cases, the 737 does not automatically trim for you when in manual flight.

This is not an apples to apples comparison. The airbus is ALWAYS in trim, by definition.
 
You must log in or register to reply here.
Back
Top