Proposed Advisory Circular For Stall And Stick Pusher Training

[ppragman]

Well, that's exactly what I was trying to beat around the bush at. Lower the nose, add power (if you can) solve the problem. Basic stuff. The AF crew really really dropped the ball on that one, but the point is, had they reached way back, then they might have thought, "hey, the airplane is descending and I'm out of power, and nothing is happening, maybe I should lower the nose." Airspeed is life.

 

[SteveC]

Air France - no valid airspeed indications and probably turbulent conditions. Somehow through the initial confusion of erroneous/missing information they end up fully stalled. The confusing thing though is that they end up in a configuration that would "normally" be equated with flying, not stalling - noise slightly high, but not abnormally so, and they have engines powered up. This is probably a configuration that would work to take them out of an impending stall, but it's not enough to take them out of a full stall. So at this point it may be very difficult for them to sort out what their actual status is.

One indication that doesn't match is a VERY high descent rate, but with no valid airspeed indication they might have discounted that as erroneous information. One of the kickers is that after the probes thawed out and could provide valid numbers, their airspeed was so low that it wouldn't display. They tried lowering the nose, but when they did so the airspeed started to come up into a valid range that the computers would accept, so they immediately sounded an alarm (horn? I forget...). I'm only guessing here, but I suspect that one thought that the pilot(s) had was that when they lowered the nose they were getting an over speed alarm, not a stall alarm, and when they respond by pulling the nose back up the alarm went away (airspeed drops back into an invalid range) which would tend to corroborate an over speed. Push nose down, alarm goes off, pull nose up, alarm goes away.

Like I said before, figuring out that you're IN a stall is the first step to recovering from one...

 

[ppragman]

If you're falling out of the sky at a high rate of speed, and you're not in a nose dive, lower the nose and add power. I can't think of a situation other than a microburst (which isn't exactly applicable to the AF447 accident) where that wouldn't be applicable. The hell with the warnings and bells, because in every plane those will lie to you, they should have had a decent idea of what their groundspeed would be and they should be able to reference that off of INS, or GPS, etc (not sure if the Bus Guys can even look at that, don't know anything about their systems other than the most basic basics). If your normal cruise speed is 400kts groundspeed, and you're getting a high sink rate, all sorts of goofy warnings, and its showing 100kts when you should be booking it? Now, its easy to monday morning quarterback this sort of thing, however, if one thing doesn't work, do something else.

This is from wiki.

10 seconds later, the plane's recorded airspeed dropped sharply from 275 knots to 60 knots. The plane's angle of attack increased, and the plane started to climb. The left-side instruments then recorded a sharp rise in airspeed to 215 knots.

A "sharp rise" in anything is cause to start comparing information from alternate sources. Airspeed sources that don't compare, or don't match with groundspeed readouts and erradic autopilot action require thought.
...Stuff...

From there until the end of the flight, the angle of attack never dropped below 35 degrees. During the last minutes, the thrust levers were in the "idle" detent position. The engines were always working, and responsive to commands.
The recordings stopped at 4 hours 14 minutes and 28 seconds absolute time (02:14:28 UTC), or 3 hours 45 minutes after takeoff. At that point, the plane's ground speed was 107 knots, and it was descending at 10,912 feet per minute, with the engines' N1's at 55%. Its pitch was 16.2 degrees (nose up), with a roll angle of 5.3 degrees left. During its descent, the plane had turned more than 180 degrees to the right to a compass heading of 270 degrees. The plane was stalled during its entire 3 minute 30 second descent from 38,000 feet.

Airspeed that doesn't make sense, ground speed was 107kts, and they were descending at 10,000+ fpm, and the engines were at idle, and the pitch attitude was 16.2 degrees nose up and the airplane was in a descending turn? They had 3 minutes and 30 seconds to recognize the stall. Ample time to recognize and safe the day, at 10,000fpm, the radar altimeter should have come on, they had so many warnings. When things aren't working to elliviate a situation, do something different.

 

[TUCKnTRUCK]

This always boggles my mind, if you can't get the plane to fly at your current power settings, change it. To hell with normal cruise power. Example, The 737 that flew into the key bridge in dc, why didn't somebody go , aww hell if we are going to hit the bridge anyways, I'm not worried about burning up the motors... And firewall it. Why didn't air France push the power all the way up at some point?

 

[dasleben]

There's more than one transport jet in the ocean right now due to unreliable airspeed indications. It can catch anyone off guard, particularly if you're sitting in cruise over the ocean on autopilot eating dinner. Instead of "It'll never be me," it should be "What if it was me?" Consider your response in the shoes of someone who gets 1-2 landings and maybe 20 minutes of stick time per month.

 

[Pilotforhire587]

This always boggles my mind, if you can't get the plane to fly at your current power settings, change it. To hell with normal cruise power. Example, The 737 that flew into the key bridge in dc, why didn't somebody go , aww hell if we are going to hit the bridge anyways, I'm not worried about burning up the motors... And firewall it. Why didn't air France push the power all the way up at some point?

I question the same thing. When we do wind-shear recovery in the sim they will occasionally give us severe if its a loft, the only way to recover is to firewall it. You don't have any other choice. I am glad they do it too. We train so that things are beaten so far into our brains we almost don't have to think about what to do in an emergency, then they throw this at us and you have to be a pilot again, too hell with the limitations, save your ass.

 

[SteveC]

I'm thinking some of you guys haven't spent much time in Level D sims trying to sort out conflicting information from multiple systems. It is much easier than you'd believe to get your mind convinced of one thing, especially when SOME of the information corroborates your belief, and it then becomes incredibly difficult to reorient yourself to a different belief.

Or maybe those pilots were just that incompetent. *shrug*

 

[jtrain609]

I'll see your shrug, and raise you a "dramatic internets."

The more I fly, the more I realize that the majority of my coworkers are, in fact, competent, and sometimes you just wind up in a bad situation. The guys on the Air France flight likely fall into that category.

 

[seagull]

When you are in a situation where most, but not all, of the things you are seeing resemble something you have seen before, the common response is to use what worked before in a similar situation. Factor in that your span of attention sharply narrows under stress, and the human traits of selecting information that supports their analysis, plus the startle response, and you have AF447 or AA903. The only way around it is to train for that scenario. We do V1 cuts well because we train for the startle response. Not true for the AF situation.

 

[Lee D]

I'm thinking some of you guys haven't spent much time in Level D sims trying to sort out conflicting information from multiple systems. It is much easier than you'd believe to get your mind convinced of one thing, especially when SOME of the information corroborates your belief, and it then becomes incredibly difficult to reorient yourself to a different belief.

*

I will second that! Once your mind gets wrapped around a particular idea or thought during a uber high stress situation like that, (Air France) it can be difficult to think beyond the idea and mentally process the other symptoms and see the big picture.

 

[FlyingPoke]

I'm about a page behind... but ah hell, who cares.

As an SIC at a 142 training facility, I am pretty familiar with the training and recovery that is being suggested by this AC.

The Stick Pusher training is done at both low and high altitude. Depending on the aircraft and wing load/shape, I've seen it take over 5,000ft to recover from the Pusher at a high altitude, and depending on the reaction of the client, 200ft- ground impact on the low altitude. As far as I know, this has been added to the curriculum and has been commonplace since the Buffalo crash findings. This realm of training is pretty interesting to watch and almost everyone (yes, even the guys with the all-important C172 time, pushing the nose down to recover from a stall) comes out happy they were shown the recovery process. Essentially, after reaching the pusher, the aircraft pre-predicts reaching the critical AOA after the slightest pull on the stick, thus inducing another pusher... which happens over, and over, and over, and over again if you dont know to hold the nose down for literally a couple thousand feet. Add in some ice for realism and an aircraft with centerline thrust issues to begin with, and things get real interesting.

I do agree though, as do some instructors around here, that the requirement for 'minimum loss of altitude' in the stall series training is getting a bit too strict and a slight loss should be acceptable to avoid over-training/over-emphasis.

 

[drunkenbeagle]

Air France - no valid airspeed indications and probably turbulent conditions. Somehow through the initial confusion of erroneous/missing information they end up fully stalled. The confusing thing though is that they end up in a configuration that would "normally" be equated with flying, not stalling - noise slightly high, but not abnormally so, and they have engines powered up. This is probably a configuration that would work to take them out of an impending stall, but it's not enough to take them out of a full stall. So at this point it may be very difficult for them to sort out what their actual status is.

If most of us were in the same position, with the same background, I doubt we would have fared much better.

I have stalled a heavy jet (in a sim), and for a few reasons, I don't think it will ever resemble the real world. It is outside the envelope of what the sim was programmed to model, and will therefore be less than accurate. In my case, the sim tech turned off motion as well (outside of what the thing was built to do again, and these things are pricey to fix). And it still took about 10,000' to recover. And we ended up inverted.

There is much truth to everyone's point of view here though. In the first place, stall avoidance is the more important issue than stall recovery. And the answer to that is pitch attitude. Holding 10 deg pitch up and keeping in CLB/TOGA power, the stall wouldn't have happened in the first place. That would be true for an A330 as much as it would be for a 172.

I was curious enough to research this, and guess what? There was an FAA AD years ago calling for just that in the A330 - AIRSPEED UNRELIABLE = PITCH 10 / CLB.

 

[Polar742]

So, this thread is real disappointing from several angles.

1) The old "stall series" really wasn't a true stall in any sense of the word. Everyone on this thread, as they have pointed out, have been taught to recover by "powering out". Guess what? That's the absolute correct response to the situation you set up. The old stall series was set up with the following goals: identify the signs of an impending stall, and fly out of it. Basically, you set up slow flight as you would in a C150 in primary training and you fly out the same way. When the first indications of a stall are present (shaker, pusher, horns, sirens or whatnot) you flew out. The goal was to prevent a full on aerostall.

2) The straightwing drivers who are tossing ideas out like they know what they are talking about are silly. Look, if it has a straight wing, whether is a 0 engine glider or a 4 Engine -7, you snap the nose over, unload the wing and start flying with a minimum of altitude loss.

Sweptwings, contrary to what some have written, are different. You stuff the yoke over on a sweptwing and the dirt comes at a rapid pace. It's just the way the airflows. You get a sweptwing in a real stall, and you'll quickly get a deep stall, then it's (probably) game over.

3) High performance airplanes have all manner of things like pushers, shakers, ventral fins, stall fences and other odd things sticking off the airframe or airfoil since the machine tends to do bad things in aerostalls.

To the best of my knowledge, I have not stalled any twin. I've flown approach to stalls like everyone else. Would I do it in a sim? Sure. In a real airplane? Not on purpose!

 

[CoffeeIcePapers]

2) The straightwing drivers who are tossing ideas out like they know what they are talking about are silly. Look, if it has a straight wing, whether is a 0 engine glider or a 4 Engine -7, you snap the nose over, unload the wing and start flying with a minimum of altitude loss.

Sweptwings, contrary to what some have written, are different. You stuff the yoke over on a sweptwing and the dirt comes at a rapid pace. It's just the way the airflows. You get a sweptwing in a real stall, and you'll quickly get a deep stall, then it's (probably) game over.

I think they are trying to prevent things like this happening:

The PF called "set max power" at 18.08:13. Perhaps confusing to the crew was the fact not only that the stall occurred at a higher than expected airspeed, but also that the stick shaker failed to activate. For the next 8 seconds, the PF continued to hold the nose up, maintaining a relatively constant pitch attitude. Popping sounds were also heard coming from one or more of the engines, and engine indications of surging were present, which can happen when airflow into the engine intakes is at an excessive angle. Airspeed continued to decay, and the aircraft began descending as the stall progressed. At 18.08:30 the PNF (pilot not flying) stated "You can take a little altitude down…" He was implying to the PF to push forward on the yoke. But at 18.08:42, he added, "Start bringing the nose back up." For the next 56 seconds, the DC-8 continued descending and began a series of roll reversals.

The most prudent thing to do here would be to push forward on the yoke to reduce the angle-of-attack. Perhaps the pilot failed to do this, because it was not necessary to do so in his previous training, since the stick shaker kept the aircraft below critical angle-of-attack

http://www.apstraining.com/2009/lessons-learned-dc-8-mishap-on-22-december-1996/

 

[TwoTwoLeft]

If aerodynamic stall recovery was taught initially from the standpoint of energy management rather than a procedural event outlined in the PTS, the logical recovery might translate much better into larger aircraft. Aerodynamic stall recovery in its absolute basic form is to ALWAYS reduce AOA. The purpose of power is to mitigate altitude loss. So the question becomes, How do you safely and affectively apply power after an aerodynamic stall recovery for the airframe you're flying and what are the control pressures required? Maintaing directional control is just as important as pitch. How does your airframe handle large, sudden applications of power at low (it's relative) airspeeds? Pilots who fly a C-150 to an A-36 to a Navajo to a Citation X, to an Extra 300 to TBM 850 are all going to have different stories to tell about their stall experiences (actual or simulated). The problem with one airframe to the next is for the pilot to be able to identify the approach of the aerodynamic stall. To over-simplify, some wings will load up and break rapidly while others will progressively mush until a deep stall is entered

 

[dasleben]

So, this thread is real disappointing from several angles.

1) The old "stall series" really wasn't a true stall in any sense of the word. Everyone on this thread, as they have pointed out, have been taught to recover by "powering out". Guess what? That's the absolute correct response to the situation you set up. The old stall series was set up with the following goals: identify the signs of an impending stall, and fly out of it. Basically, you set up slow flight as you would in a C150 in primary training and you fly out the same way. When the first indications of a stall are present (shaker, pusher, horns, sirens or whatnot) you flew out. The goal was to prevent a full on aerostall.

2) The straightwing drivers who are tossing ideas out like they know what they are talking about are silly. Look, if it has a straight wing, whether is a 0 engine glider or a 4 Engine -7, you snap the nose over, unload the wing and start flying with a minimum of altitude loss.

Sweptwings, contrary to what some have written, are different. You stuff the yoke over on a sweptwing and the dirt comes at a rapid pace. It's just the way the airflows. You get a sweptwing in a real stall, and you'll quickly get a deep stall, then it's (probably) game over.

3) High performance airplanes have all manner of things like pushers, shakers, ventral fins, stall fences and other odd things sticking off the airframe or airfoil since the machine tends to do bad things in aerostalls.

To the best of my knowledge, I have not stalled any twin. I've flown approach to stalls like everyone else. Would I do it in a sim? Sure. In a real airplane? Not on purpose!

Solid.

On one of my last recurrent sims, we did a high-altitude stall in the 767 at a normal cruise weight for the mid-30s (somewhere in the range of 350-360,000 lbs). We lost at least 5,000 feet, possibly closer to 10,000. That was predicated on me:

1) Knowing what was going on ahead of time, i.e. no startle factor or trying to straighten out conflicting information, and

2) Consciously trying to minimize altitude loss at max thrust while reducing the AoA.

I hate to say it, but if Air France got down to 60 knots (wow!) during the descent, they would have still been in serious trouble had they figured out the problem as high as in the 20s.

 

[seagull]

If aerodynamic stall recovery was taught initially from the standpoint of energy management rather than a procedural event outlined in the PTS, the logical recovery might translate much better into larger aircraft. Aerodynamic stall recovery in its absolute basic form is to ALWAYS reduce AOA. The purpose of power is to mitigate altitude loss. So the question becomes, How do you safely and affectively apply power after an aerodynamic stall recovery for the airframe you're flying and what are the control pressures required? Maintaing directional control is just as important as pitch. How does your airframe handle large, sudden applications of power at low (it's relative) airspeeds? Pilots who fly a C-150 to an A-36 to a Navajo to a Citation X, to an Extra 300 to TBM 850 are all going to have different stories to tell about their stall experiences (actual or simulated). The problem with one airframe to the next is for the pilot to be able to identify the approach of the aerodynamic stall. To over-simplify, some wings will load up and break rapidly while others will progressively mush until a deep stall is entered

Thrust for a jet engine is considerably more complex http://home.anadolu.edu.tr/~mcavcar/common/Jetengine.pdf and the responses of swept wings (flat lift curve slopes) are also quite a bit different (as has previously been explained). You really cannot correlate your experience flying a light, straight winged, prop aircraft with the associated wind noise, sitting near the CG and simple systems for instrumentation and flight controls to what the experience is when flying an advance design large transport aircraft.

 

[TwoTwoLeft]

Thrust for a jet engine is considerably more complex http://home.anadolu.edu.tr/~mcavcar/common/Jetengine.pdf and the responses of swept wings (flat lift curve slopes) are also quite a bit different (as has previously been explained). You really cannot correlate your experience flying a light, straight winged, prop aircraft with the associated wind noise, sitting near the CG and simple systems for instrumentation and flight controls to what the experience is when flying an advance design large transport aircraft.

Not really sure of purpose of this post.... But thanks for the link I guess?

 

[nosehair]

WoW! I am dumbfounded! Seriously!

All these years I've been flying and teaching small prop airplanes, I have always assumed that big expensive aircraft all had AoA indicators. How can this be?

 

[derg]

Knowing the FAA and fixing things this week, they're going to improve "stall training" by requiring all pilots to conduct "lomcevaks", inverted, in IMC with a monkey suit on.