Air France almost loses another one

SkyHog

SkyHog

Touchdown! Greaser!
Joined
Feb 23, 2005
Messages
18,433
Location
Castle Rock, CO
Display Name
Display name:
Everything Offends Me
http://www.dailymail.co.uk/news/art...topilot-fails-drama-echoing-Brazil-crash.html

An Air France jet was just seconds from nose-diving to disaster when the autopilot failed during extreme turbulence.
The mechanism malfunctioned as the plane hit a storm at 35,000ft.
The Airbus, with up to 375 people on board, then climbed sharply to 38,000ft losing speed as it did so.
Its speed dropped to just three knots from stalling.
Click to expand...

Are we looking at a potential issue now, or is this another one-off?
 
Last edited:
One-off; the autopilot disconnected, the plane zoomed, the pilots corrected, life went on.
 
Three knots from stalling, yawn. Then, the pilots would have had to lower the nose a bit or add power to recover. Not saying there was nothing to it, but not a lot.

Best,

Dave
 
The autopilot is supposed to disconnect during "extreme turbulence", even during turbulence that is less than extreme. Remember it's not such a great idea to try to maintain altitude in those conditions.
 
Dave Siciliano said:
Three knots from stalling, yawn. Then, the pilots would have had to lower the nose a bit or add power to recover. Not saying there was nothing to it, but not a lot.

Best,

Dave
Click to expand...


We know that air france pilots have issues with stall recovery
 
Everskyward said:
The autopilot is supposed to disconnect during "extreme turbulence", even during turbulence that is less than extreme. Remember it's not such a great idea to try to maintain altitude in those conditions.
Click to expand...

Based on some of the discussions over at PPW, it sounds like this was a case of the 'bus doing exactly what it was designed to do - protecting itself from an overspeed condition. Apparently, it is a protective feature of the aircraft that will disconnect the A/P and command a pitch up to reduce speed. Even if the pilot applies forward stick pressure, the computer will over-ride the input and maintain the attitude until it slows.

Here is the excerpt posted from the aircraft manual:
High Speed Protection

High speed protection prevents aircraft damage due to excessive speed. High speed
protection activates slightly above VMO/MMO, at the speed depicted by a green equal
sign.

When high speed protection activates:

• The autopilot(s) disconnect.

• Nose down pitch trim is inhibited.

• The flight control computers attempt to limit further acceleration by commanding a gentle pitch up. This protection limits maximum airspeed by automatically increasing the aircraft’s pitch attitude even if the sidestick is held full forward.

• The bank angle limit is reduced from 67° to 45° (limit marks do not move).

• Positive spiral stability becomes active at all bank angles (instead of the usual 33° in normal law). Therefore, the sidestick must be held to maintain any bank angle. If the sidestick is released the aircraft returns to a bank angle of 0°.

• The CRC and MASTER WARN lights activate and an O/SPEED (overspeed) warning is displayed on the E/WD. The CRC can only be silenced by slowing down or pushing the EMER CANC pb. Holding the sidestick full forward will allow some exceedance of the limit speed (this permits limited maneuvering if necessary). The amount of forward sidestick held and
the rate at which the sidestick was moved forward affects how far the aircraft’s speed will increase above VMO/MMO. If the pilot holds the sidestick full forward, the speed will stabilize above VMO/MMO. When the sidestick is returned to neutral, the speed will return to VMO/MMO. High speed protection is deactivated when the aircraft decreases below VMO/MMO.
Click to expand...

So, the pitch up and climb is part of the design. The big question is why it got so bloody slow. Possibly the pilots not reacting quicky enough or not understanding what the aircraft was doing.

Personally, I don't think I like the idea of flying a plane that thinks for itself.
 
dell30rb said:
We know that air france pilots have issues with stall recovery
Click to expand...
I don't think that's a fair statement at all.

Fearless Tower said:
Based on some of the discussions over at PPW, it sounds like this was a case of the 'bus doing exactly what it was designed to do - protecting itself from an overspeed condition. Apparently, it is a protective feature of the aircraft that will disconnect the A/P and command a pitch up to reduce speed. Even if the pilot applies forward stick pressure, the computer will over-ride the input and maintain the attitude until it slows.

Here is the excerpt posted from the aircraft manual:


So, the pitch up and climb is part of the design. The big question is why it got so bloody slow. Possibly the pilots not reacting quicky enough or not understanding what the aircraft was doing.

Personally, I don't think I like the idea of flying a plane that thinks for itself.
Click to expand...
That's not entirely uncommon of a design. Our newer aircraft (the glass panel ones) have a similar protection. If you're on a/p and in either VS mode (you directly select the target FPM down and it holds it), or VPath mode (you give it a crossing restriction and it descends for it based on the target slope or target vert speed you give it) and you let it get more than 2kts over the high speed cue, it starts beeping and flashing like mad, and switches to IAS/Mach hold and dials down to Vmo-2 or Mmo-.05 and stays there until you command something else (and are no longer overspeeding). Depending upon how far behind the plane you get and how fast you're going down, that can easily result in a level off or climb.

Every plane I've ever flown, the autopilot will kick off in a good bit of turbulence. As Mari pointed out, that's a very good design feature. Some are more sensitive than others, but they all do it.

So add that above mentioned overspeed protection with a good bit of turbulence...now the plane has commanded itself to climb AND taken away the autopilot. Potentially dicey situation? Sure, but it's better than the autopilot holding on and fighting until it causes the destruction of the plane through overspeeding or over G-ing.

As far as why this plane got so slow...just a totally WAG, but if the plane had inhibited forward stick movement AND locked out forward trim, perhaps the thing was just so far aft stick/aft trim that it took them a while to get it commanded back down to a normal deck angle?
 
Last edited:
Heck, my autopilot *creates* turbulence out of smooth air, and nothing ever kicks it off! ;)
 
Dave Siciliano said:
Three knots from stalling, yawn. Then, the pilots would have had to lower the nose a bit or add power to recover. Not saying there was nothing to it, but not a lot.

Best,

Dave
Click to expand...

Sad thing is, that should have been said about AF 447 as well...:rolleyes2:
 
Dave Siciliano said:
Three knots from stalling, yawn. Then, the pilots would have had to lower the nose a bit or add power to recover. Not saying there was nothing to it, but not a lot.

Best,

Dave
Click to expand...

Perhaps, but at that altitude you're going to lose about eight to ten thousand feet in the recovery, all while listening to the clattering of the stick shaker on the way down. Also hoping like hell that there is nobody underneath you there in the RSVM airspace. So it's gonna be a slightly different ride than when it happens in the mighty Cessna. The big question is why do these French idiots keep driving on into to turbulence ?
 
Teller1900 said:
I don't think that's a fair statement at all.
Click to expand...


Given the outcome of 447... its easy to see why this is a big deal and a possible similar situation. Hopefully they learned their lesson and i'm sure AF has had their pilots in the sim drilling on stall recovery. Not just AF but all airline pilots, and especially airbus pilots.

The autopilot kicking off while in severe turbulence... probably a very distracting environment for those pilots. They were getting shaken up like hell and i'm sure the airbus was blinking and beeping like mad. Exactly what happened to 447.
 
Last edited:
Art VanDelay said:
Perhaps, but at that altitude you're going to lose about eight to ten thousand feet in the recovery, all while listening to the clattering of the stick shaker on the way down. Also hoping like hell that there is nobody underneath you there in the RSVM airspace. So it's gonna be a slightly different ride than when it happens in the mighty Cessna. The big question is why do these French idiots keep driving on into to turbulence ?
Click to expand...

First, jets like this one add power and have a set attitude to normally recover. I don't know what you fly, but losing eight to ten thousand feet to recover is more like spin recover. If they catch the stall quickly, they'd lose little altitude. Since they had climbed when the AP kicked off, which it was supposed to do in turbulence, there was a good chance they wouldn't ever get back to the originally assigned altitude unless it was a very strong stall.

I'm not saying it wasn't noteworthy. As has been said, getting within three knots of stall isn't a good place to be. There's very little room between stall and Vne at that altitude, that's why they mostly fly AP up there. Still, the crew did what they were supposed to.

Best,

Dave
 
Something happened to Otto?

attachment.php
 

Attachments

  • otto.jpg
    otto.jpg
    9.8 KB · Views: 41
You must log in or register to reply here.
Back
Top