Linktrained,

I fully agree that some respect is called for. Though I suppose with regards to test pilot territory you're talking about unstalling a widebody going down 10000fpm at 60 deg AoA, not about keeping their plane reasonably straight and level for a minute or so? After all the former only resulted from failing at the latter...

There are various ways to tell 'non-test-pilots' that they are entering 'test pilot territory', and test pilots lost their lives to get those boundaries established.

"Stall Stall Stall" is one, and unlike what some people here still seem to think, it's a warning about 'approach to stall', it doesn't mean (yet) the aircraft is stalled.

The same applies to stick shakers and stick pushers.....

Once those warnings are not heeded (either because of the perceived or known or 'advertised' unreliabilty of the warnings itself, or the incapacity of the pilot of how to deal with them) disaster in 'test pilot territory' iis often only just around the corner.

"Those who can't learn from history, are bound to repeat it", and another test pilot, and other pilots and passengers, will have died in vain.

PA and PNF did not appear to KNOW that they were stalled OR descending at 10,000 fpm, until they were at 10,000ft. ( AoA is on the FDR, I believe.)

Perhaps Approach to Stall Training should be called 447, just to keep that lesson well learned by future generations.

More and more I keep on thinking this crew had been probably well prepared and rehearsed for take-offs and landings but cruising had been always left to the automatic system. As time went by and comfortable routines became the norm, they unknowingly became less proficient high altitude PILOTS. It didn't take much for life to prove them deadly wrong. All three of them in their minds avoided looking and addressing the fast decrease of Altitude because they were simply blank about what to do about it. They never thought that one day they might have to fly manually the airplane without the help of the FBW during the cruising phase. Alternate law or even Direct Law, impossible. Those situations could never happen because is simply too extreme and it would mean that we'll all be dead by then. But the remotely impossible and improbable extreme came to be and the tragic outcome did happened as anticipated. That final outcome could have had a different and happier ending but good old human complacency was already hard at work and too deeply rooted .

Yet there have been plenty of survivable UAS events including in Airbus's. For example see Incident: Jetstar A332 near Guam on Oct 28th 2009, series of ECAM messages
where the aircraft stays within 100 feet of its assigned FL during the UAS event.

Holes in Cheese
 

This is not an equivalent incident. The other "holes in the cheese" were not there - in moderate turbulence at night and automatics handing over not in straight-and-level

I am curious...are there any mechanical attitude instruments in the Airbus or is it all tube displays.

I ponder a cockpit where all the tv screens go out, lighting strike...what would the pilot have for attitude reference?

Hi,

Indeed .. you right
In the other cases .. it was good pilots in the cockpit
It was not the case for AF447 and that's was a big supplementary hole in the cheese (read BEA reports and the BEA press meetings)
That's the sad reality .. and this is a real problem for Air France

That's a question I've asked myself a few times, too.
The present-day standby instruments use small LCD screens too, which still rely on some kind of separate secure electrical supply, and presumably so does the attitude reference (gyro of some type).

I 'grew up' in the days of the "SFENA", a small 3" standby horizon with its own hi-speed hi-inertia gyro, that would run for several minutes after a total loss of power. The display was purely mechanical, so the attitude information was still displayed.
At some point it equipped about 70% of the world's airliners, including Concorde and 747.
And it's credited with being the ultimate aid in a few OMG situations, and bringing the plane home (a Caravelle with a total electrical failure - or lightning strike, I'm not sure - comes to mind).


I worked for SFENA, but not in that division. I saw the first LCD display standby horizons arrive, but never got the answers to their reliabilty compared to the old "spinning top" version in extreme situations.

Test Pilot
 

A few misguided comments about test pilot territory.
From what I have heard liner test pilots say when they test the edge of the envelope it is in a highly controlled and low energy way. The goal is to keep control and have the ability to return to the envelope quickly.
These guys pushed aggressively through the envelope to the point of invalidating (breaking) the stall warning. The pilots took AF447 it beyond normal test conditions.
AF need to address why all 3 pilots with their training would not mention "stall" or "décrochage". All 3 pilots could hear it. I don't think Airbus and BEA would have that answer.

Now we'll get somewhere:

Airplanes are so automatic, pilots are forgetting how to fly them, says federal panel - NYPOST.com

This won't fix anything because once or twice a year having upset recovery does not a pilot make. They will check one more square for the ten minutes they did that training but it won't fix anything because ten minutes is not even close to enough time and the airlines don't want to invest any more sim time in a meaninful two hr session. There are times during flight you can handfly legally per FAA usually now in the lower altitudes below RVSM but so many airlines don't want their pilots to do it at all. On our major airline almost everybody did a lot of hand flying so we didn't have a big problem with automation dependency. Also all had been on board at least 12 years so no new guys.

So, as I've been saying from the outset, pilots' unions and airline safety campaign groups need to band together and, as FDR told the progressive groups that helped him into power, "make [them] do it".

As I said earlier, you're probably not going to get a better time than this for the next decade, or at least until the next LOC accident claims more lives. Ask the right questions and tell your stories to the right people and you could have airline management on the ropes.

The Jetstar incident was nighttime, around midnight local time, in cloud and with lightning around. But this is just one example. I believe there is a list somewhere of over 30 UAS incidents that were successfully recovered from.

Bubbers44

I went thru the "old" school USAF UPT, lots of acro, lots of under the hood acro, and lots of recoveries; then fighter RTU same torture by ex-Vietnam types who loved handing the plane back, under the "tent", out of speed, straight up to see how you would handle it. Now, even in the AF, the T-1 program has none of that. Funny, when I last went to NTPS and did upset training, I all came back like riding a bicycle despite years of not doing it.

The lesson: this training needs o be "imprinted" early on and reinforced to the point of reaction not deep thinking.

In other words: Has some phase of the training de-sensitized the crews from hearing and responding to the stall alarm? Do Sim instructors debrief items while the stall alarm is sounding? Have crews somehow learned to ignore or 'listen through' the stall alarm?

The rudder pedal shaker in the F-4 was excellent for gaining my attention promptly. Several threads before, I mentioned a seat shaker as probably the most viable tactile stall warning technique on a stick flown FBW transport aircraft where the crew may fly with feet off pedals and even hands off stick. With the AF447 crew demonstrating that they did not respect/respond to the stall warning annunciation and crickets, there must be a need for a better way to get a pilot's attention. (Yes, PF did appy TOGA at the second set of stall alarms together with aft stick to keep the nose from falling :*)

Agree. Tactile cues used to be thought essential in such a role. No need to introduce seat shakers though. Stick shakers are a tried and tested way of attracting a crews attention. But they aren't part of the AB FBW philosophy and it might be quite a wrench to introduce them now.

With all of their stall warnings going off and not responding, I don't think a baseball bat hitting them on the head would have helped. The captain by the time he got up there in his groggy condition and as confusing as the situation was by then probably couldn't help in the limited time he had to respond. Maybe Air France should hire qualified pilots in the future. Maybe everybody should. I know, that costs more money.

If I ever need heart surgery or brain surgery, I hope the airline industry isn't hiring my doctor.

As far as i know both Boeing and Airbus use a Thales system as standby instrument. In the 737 the ISFD (integrated standby flight display) has its own internal battery pack that provides power for 150 minutes completely independent of the rest of the aircraft, its own internal inertial sensors which even act as the third inertial platform during fail operational auto land operation. The only inputs it receives is heading information from the left IRS and ILS information from NAV 1.

According to our maintenance those units have a much lower failure rate than their old mechanical counterparts, not surprising as they have no moving parts. Additionally those old units were powered by the general aircraft batteries which only provide enough power for 60 minutes. Inertia keeps them running for a bit after power is off, however the attitude reference becomes unreliable very soon after power is removed, unlike the ISFD which runs happily on its own even after the rest of the instruments fail because the aircrafts battery is depleted. In that case you need a flashlight to read the magnetic compass for heading reference though. Additionally it uses the same presentation as the PFDs, has no parallax error and is therefore easy to use from both seats, and has its own light-source of course which keeps it usable for those 150 minutes in darkness as well.

Mad Scientist

Do you really think you can shake the short Airbus stick sufficiently to draw attention without also making control inputs?
And what if the pilot is not holding onto the stick when the alarm is triggered?
Then again, how would you alert the PNF?

It is relatively easy to shake the seat and no chance of unintended control input. It will get your attention. And it doesn't occupy your overworked visual-aural channel.

It isn't part of AB FBW philosophy because the aircraft were supposedly virtually unstallable. Guess that bit of philosophy needs a revisit in light of current experience:}.

One for the vultures ...
 

Le Figaro - France : L'incident qui relance le dbat sur l'AF 447 du vol Rio-Paris

Le Figaro today publishes a report of a near-perfect repeat of the AF447 incident - but without the 200+ fatalities.

Assuming the incident and the document are real, and assuming the much-criticised original pitot tubes were replaced as recommended, this would seem to narrow the debate over potential causes by one parameter at least.

I must disagree, Sir.

- no speed data loss on this A340, or so it seems so far
- on the other hand, no overspeed alarm on AF447 (which would explain the climb) (and I know that the CVR mentionned "je crois qu'on a une vitesse de fou, là" or something like that, but it was much later, far into the stall).
- no mention of a reversion to alternate on the A340 from Caracas. Assuming the NORMAL LAW, the climb seems adequate, then. If overspeed, then reduce thrust and climb, that's the way the high speed protection works AFAIK.
- if still in NORMAL LAW, the stall was not to be feared on this A340 : the overspeed protection worked, I see no reason that the high AoA protection would not have kicked in, if needed (perhaps it has, for all we know).


NB: This new incident should (and is, apparently) investigated. This is not a subject of debate for me. :ok:
But the article seems to link AF447 & this incident far too much. := Paper selling agenda? :oh:

Me again (sorry)

What seems, for now, more linked to AF447 is the publication yesterday (sept the 5, 2011) by the BEA of the report of the work group on the triggered transmission of flight data (inflight).
This report appears to be available only in french, so far. There.
I don't know if an english version will follow?

To help our non french-speaking friends, here is a quick translation if needed.
It is an Air France Flight safety information report.

* Level bust en turbulences
En croisière au FL 350, en IMC de nuit, un A340 en provenance de CCS traverse une zone de
Cruising at night at FL350 in IMC, an airbus 340 coming from Caracas enters a zone of severe
turbulences sévères, 10 nm au nord de la position 18N060W.
turbulence ; 10 nautical north of 18N060W
L'avion subit un fort gradient de vent (+22 kt de vent effectif en 5 s), avec une vitesse atteignant
The airplane meets a strong wind shear (+ 22kts in 5 seconds) with an increase of speed up to Mach
Mach 0.88 (MMO de 0.86) pour une CAS=304 kt avec déclenchement de l'alarme Overspeed.
0,88 (MMO is 0,86) with a CAS of 304kt triggering the Overspeed alarm.
Simultanément, une déconnexion de l'AP se produit. L'assiette de l'avion augmente ensuite
At that moment, the Auto pilot disconnects. The plane’s pitch attitude reaches progressively
progressivement vers 11°, avec un vario voisin de 5000 ft/min.
11 degrees with a climb rate close to 5000 fett per minute.
Moins d'une minute après la déconnexion de l'AP, l'avion atteint le FL380 avec une vitesse en
Less than one minute after the Auto pilot has disconnected the plane reaches FL 380
diminution vers Mach 0.66 pour une CAS=205 kt (Mach GD =0.78, GD=276 kt).
With a speed coming down to M 0,66 and a CAS of 205 knots
La diminution de l'intensité des turbulences s'accompagne d'une réduction de l'assiette de l'avion,
The lessening of the turbulence intensity is met by a reduction of the pitch angle ;
puis d'un retour à des paramètres de vol normaux.
Then by a return to normal flight parameters.
La période de turbulences dure deux minutes avec une intensité plus prononcée lors des 30
The length of the turbulence event is of two minutes with a peak during the first thirty seconds.
premières secondes. Plusieurs pics d'accélération verticale sont enregistrés, les valeurs limites
Several peaks of vertical accelaration are recorded with peaks at 1,7g and 0,5 g.
atteignant +1.7 g/+0.5 g.
Les informations météorologiques fournies à l'équipage n'indiquaient pas de turbulences dans cette zone du parcours. Le radar météo était en fonctionnement.
Meteorological information in the flight crew’s briefing did not indicate any turbulence on this part of their route.The weather radar was on.
Aucun blessé n'est à déplorer parmi les passagers et 2 CRAT bénins sont rédigés par l'équipage commercial.
No passenger was harmed and two medical reports were filed by the cabin crew for not serious injuries.
La rapidité de transmission de l'annonce «ici le poste de pilotage, PNC assis attachés, turbulences fortes », a contribué à limiter les dommages corporels en cabine.
The speed with which the cockpit alerted the cabin (here is the cockpit speaking, commercial crew sit down and fasten your belts ; severe turbulence coming » helped to contain physical damage to the crew.
Mise à jour 28/07/2011 : cet évènement fera l'objet d'une enquête du BEA à laquelle Air France sera associée.
This event will be the subject of a BEA enquiry and report, Air France will cooperate in that work.


Forgive my sometimes stumbling english. Te protection associated in normal law with an overspeed did in fact apply.
We lack information on angle of attack reached, and engine's thrust variations. I think we should wait for the BEA reportt. But i cannot help thinking that "the cure is sometimes worse than the illness"...Heavy handed automatic protection at high altitude maight have led to a very difficult situation (the speed went down to 205 knots after all) compared to a momentary excursion in the overspeed, which would not last as it was due to turbulence in level flight).

As far as I can gather with my rusty French, in this case there was no UAS and no alternate law, but rather overspeed protection kicking in? If so this would have been quite different from 447...

There seem to be some similarities with this incident.

Hi,

http://www.pprune.org/tech-log/46062...ml#post6683666

I wonder if this is the result of the automatism (overspeed protection in normal law) or if this is the result of the automatism mixed with a pilot action (pull on SS)
Nothing about in the note ...
Curious to read BEA preliminary report about this (provided the recorder FDR was preserved and not erased like for the AF445 case !)

The BEA's Report by the Triggered Transmission of Flight Data Group was published last March and I recall has been available on their website since about 20 March 2011.

Edit :: the correct date is:- 15 June 2011.

Control law sub-mode
 

I had seen the reports about the A340 incident, but didn't see the control law logic back then. I subsequently saw that "sub-law" in some of the great data we have had on this forum.

I initially considered the alpha prot limit as a contributing cause, especially the "automatic" pull to that limit despite the basic gee command. then we got the pilot stick inputs, so I disregarded it.

One mo'time! Confusing reversion scheme and flying so close to the "limits" programmed that we get "overspeed", "alpha prot" and "stall warnings" when the jet deviates only slightly from a "stable" flight condition.

Big surprise about the recommendations was lack of a recommendation to look at the flight control reversion sequence and logic and implementation. Although the pilot attempted to re-engage the A/P twice ( wow! what?), he took control and recovered.

@ mm43 : oops, my mistake, then. The english version was first, for once.
Anyway, thanks for pointing it :)

Surprise...shock...horror. And not just APs! Engines, radios, NAVs are also allowed to malfunction from time to time and poor pilots are expected to cope with it! Is it realistic thing to expect from mere human being?

Nope, they are not to be considered poor piloting compensators. They are fitted if certifying authority deems that aeroplane stalling characteristics warrant their installation. They may help slightly distracted pilot but they are of limited value with totally incapacitated one. As is anything we came up with so far.

Aeroplane was not getting away from them. CM2 has chased her away pretty actively.

Is it reasonable to ask me to control the roll by spoilers alone when right roll control linkage gets jammed and I get no feedback via yoke whatsoever? Is it reasonable to expect my F/O to fly us back to the tera firma with no spoilers and thus reduced control if my side is stuck? Was it reasonable to expect Genotte/Michielsen to survive when Gremlin hit took out all of their A300 hydraulics? How about Fitch/Haynek/Records?

For Finnegan's sake, the pilots are expected to cope with the degradation of flying qualities as far as they are able, not by the law of the men, but by the simple law of self-preservation.

Absolutely correct. Today's training is what aviation authorities believe to be optimal trade-off between training value, risks associated with flying and costs.


1. Yes, I do, provided it turns out FBW Airbi meet criteria for their installation. It's stickshaker, not stickwaver.
2. Aural warning, already patented.
3. See 2.

In general terms, I wouldn't say so. I don't think that background checks will find much extraordinary, like more than usual amount of re-training. IMHO there was something that debilitated them at the moment probes froze and did not let them out of its grip till the watery end. This something is what has do be drawn out into the light. To paraphrase Arab proverb: they were like children going into lion's den, unaware that the old nasty lion lives at high AoA. They forgot an old aviation cliche:

Hi Clandestino,

In Post #751, I mentioned 3 QRH procedures which specify the use of Full Back Stick. It is not mentioned anywhere that the "piloting technique" of respecting Alpha Max should be employed, and to relax the back pressure / reduce the angle of attack when very close.

If in Normal Law, the piloting technique of respecting Alpha Max is not trained, then there is a chance that some crews won't respect it in Alternate Law either.

Hi Clandestino, I see you are an equal opportunity commentor.
My experience with the rudder pedal shaker was that it was pretty vigorous and if such a shaker were to be bolted into a FBW control system, it would probably cause excessive motion. But the hand is much more sensitive than the foot, so you may well be correct.

Only problem I see with your comment 2 & 3 is that it doesn't seem to have worked in real life when the chips were down.

But I'm suspecting denial as being a root cause for lack of response among the 2 copilots. Something along the lines of, "How will we explain getting into a stall.?" How will we recover from it? Maybe it isn't a stall."

And the Captain was led down the path by the copilots who didn't want to tell him how badly they had lost control.
I've seen it before in aviation-sort of a better to die than to look bad attitude.

A pilot who maneuvers aircraft should be keeping a mental log of where their energy is. Something along the lines of: "I've gained 3000 feet, I must have lost 70 knots, but if I dive back down, I'll get it back."

The PNF seemed to be aware of how the PF was mis-handling the aircraft and should have been running a rudimentary energy tally. But it appears he was instead completely clueless.
Of course maybe I am expecting too much. None of the AF447 crew were ever tactical jet pilots.

You are correct but then we can agree that there are chances that some crews will take off without clearance and smash their half-airborne 747 into side of another, some will put their trust in sole failed instrument and spin their 757 into sea, some will dive MD-11 from 4000 ft so steeply to leave themselves no room for recovery... etc.

That every Airbus pilot absolutely must know how protections work, when they are to be relied upon, when they are not to be trusted and that "Airbus can not be stalled" is a simple myth, pushed forward by misinformed & ignorant, shouldn't be matter of discussion at all. We may discus why it isn't so in real life.

If chances of needing maximum climb performance, achieved by pulling the stick to backstop in normal law are 1 to x, and chances of having no protections are 1 to y, then chances of having to observe alpha max on your own, manually are 1 to x times y. Not as easy as letting the computer wring out the maximum performance but not impossibly difficult for a pilot, either. IMHO it's very small (and avoidable) chance for a pilot to fly himself into windshear or CFTT in degraded control law. BTW, there's good reason we call those electronic gizmos "computers" and not "thinkers" or "deciders".

I can pretty safely assume that pedal shaker didn't affect the handling of your Rhino at all. Shakers oscillate with high frequency and low amplitude. Because of both, controls and aeroplane inertia, they don't affect the flightpath. With ideal control harmony being 1-2-3, I believe that eventual sidestickshaker needn't be nowhere near powerful as pedalshaker.


It didn't work for the crews of AF447. According to BEA's intermediate report no2, there were 37 cases of pitot blockage on 330/340, apart from AF447. 13 of them could be studied further, as there was enough data available. In 5 of them crew descended following the stall warning.

Doesn't look like systemic problem to me.

If I include late captain Marvin Renslow's paradoxical reaction to stall warning, shaker and pusher into picture, then we might have systemic problem after all but it goes much deeper than aeroplane model, make and manufacturer.

Another puzzling aspect of the accident is that nose-up attitude and unwinding altimeters should make clear to him what is happening yet did not.

Absolutely correct and not debatable.

Of course, it is possible to do both at the same time.

Stick shakers, et al
 

I go with Clandestino about using a "vibrator" on the stick. In the AF447 case, the dude that kept pulling back would have felt it, as he was not just sitting there watching Otto try to fly the jet.

The 'bus stick is displacement, not force as we had in the Viper. The sensors can easily be tailored to pay no attention to the "buzzing" stick.

Need a Concorde dude here, as I bet its "stall" was very benign and a shaker would warn of getting too high of an AoA. My own experience in a delta was that there was zero "stall" as we think of it. The thing just got "mushy" as AoA went way up and drag exceeded power available. Very smooth, with only a very slight buzz that you could hear first, then got to where you could feel it. Viper was almost the same due to the automatic leading edge flaps.

Energy awareness discussion needs to be expanded here. Seems a basic component of the "pilot" skills that can be learned and trained to accomodate. The sailplane folks here shuld be very aware of energy management. I only had about 4 minutes of it, as Sully did, when my motors quit and I deadsticked the jet onto a runway.

Quote:

---

Of course, it is possible to do both at the same time.

---

Yes, even while sitting in an ejection seat.:(

gums,
You're right, Concorde "stalled", or rather "mushed" much like another delta.
Not a good idea, because by then you were well on the back of the power curve, so there were both a stick shaker and a stick pusher (SFC, safety flight computer) to keep you out of trouble.

A Camel is a Horse Designed by a Committee
 

And the Airbus control laws are designed by a committee of engineers over a period of several months.

I am sure that the engineers were thoroughly conscientious, but it definitely looks that the pilots did not know what airplane they were trying to fly -- well the word Alternate was pronounced, but there was no learned disquisition in the cockpit on the implications of Alt 2B, even though I am sure the committee of engineers spent several days talking it over.

Engineers and programmers can get a bit carried away with themselves when getting into recondite corners -- here we see a proliferation of flight laws and a crew unprepared for Alt 2B combined with UAS at high altitude.

Applying KISS retroactively, the pilots would have known what airplane they were flying if Normal Law was flown with a touch pad and the stick, rudders and trim wheel were reserved for Direct Law;)

In a previous thread I mentioned that the VSI and altitude displays did not alert the crew to the descent rate problem until the 10,000' digit dropped off the altitude readout -- i.e. the readout went from 5 digits to four. Only at that time did the crew verbalise the altitude.

Hi,

Information, 7 septembre 2011