Absolutely, but these guys were behind the curve from the outset - they never even recognised that they were stalled, and to me that's far more worrying than the minutae of autotrim operation.

I'm not denying it at all, but would it not be fair to say that rapid descent despite full power and nose-up attitude would be enough of a clue no matter what you're flying?

Really? I'd have thought that ensuring stability and keeping control inputs gentle until getting a feel for the situation would be fundamental to flight management.

Obviously "stay within the flight envelope" doesn't cover everything, but following that advice and using it as a baseline surely increases your chances of coming out of things in one piece?

I wasn't necessarily referring to you there. Also note that I didn't bring up A vs. B at all! :)

For which I salute you - military investigation has a reputation for suffering political pressure over and above the civilian equivalent. As I said, I wasn't referring to you in particular.

At the same time, I think that the crux of the "happenings" was a poorly-trained crew with significant CRM problems being startled into making some very poor decisions without first properly assessing the situation. It's a matter of a fraction of a second between the onset of what should have been a relatively harmless UAS anomaly (which over thirty crews had successfully resolved prior to this incident) and the PF deciding that consistent NU input is the solution to the problem:

• Published UAS procedure? Disregarded.
• Transient Stall Warning? Disregarded.
• Maximum altitude (which had been agreed with PNF only a few minutes prior)? Disregarded.

All this at the point where the aircraft has barely begun the zoom climb. They're already in danger and autotrim hasn't even come into it.

The reason I find all this back-and-forth about autotrim (specifically the suggestion that it hindered recovery) frustrating is that it ignores the salient point that autotrim only worked against the PF because he was doing the wrong thing. If the PF had recognised the stall and started a recovery with full nose-down on the sidestick, autotrim would actually have *assisted* in the recovery.

I think it's present at the start of an airline career regardless of type - what bothers me is the development of complacency over time.

I think that's a little strong - as you said, there are several posters who do reflexively bash Airbus, and (certainly coming from me) it's only those with a track record of doing so that have been called on it.

Not at all - in fact what tends to happen is that a post will be hijacked by one of the known agenda-holders, for which they will be pulled up. To the best of my knowledge I don't think anyone neutral on the matter has been singled out.


I don't think there's a single all-encompassing defender of Airbus systems on these boards at all (myself included). I'll point out if someone's repeating falsehoods or being less than honest, but never any more than that.

[EDIT : And it cuts both ways - if any of you encounter, either in training or on the line, someone who states that it's impossible to stall a FBW Airbus under any circumstances - frickin' report them! What they're saying is not merely wrong, it's dangerously wrong!

One last thing - as an experienced techie, I'll get asked for advice from time to time. If the subject of Apple comes up, then I'll automatically preface any response I give with the fact that I utterly loathe the Apple Corporation and its products for personal and professional reasons - and that as such my advice will be coloured by that fact. I do this because as someone whose advice is considered worth listening to in that field, I have a responsibility to be honest, and were I to dissuade a non-techie - who would nevertheless be a good fit for Apple - based on that prejudice alone, I'd be doing both them and my own reputation a disservice.]

You are saying, that A320 in Alternate LAW has always low speed protection available? AF447 was in ALT2B, communicated by the system as ALT LAW, and no protections left except load factor protection.

Please explain.

Wise words, but this was not the only crew in the past being part of a LOC accident. There is a training problem and a interface problem. Therefore BEA recomends the availability of AOA indication.


I´m not in the picture to what post you are referring. I never would object that statement.

Staying away from work would do the same. Just skip the crap with "staying in the flightenvelope " and replace it with the above quote, and we do not need to debate about it.


Are you talking about third world country?

Nothing new on the planet. In the analysis of BEA not one single crew was doing the complete UAS procedure.
There was no transient stall warning. It was a real warning, which should have lastetd longer, but due to the extreme low false airspeed and therefore shift of trigger boundary it only sounded that short.

Unfortunately you prooved it yourself, that it can get confusing, even when sitting in an armchair and not in the pointy end at FL350 over the ITC. And you made my other point, that some rely too much on protections without knowing what kind protections are available under what kind of law. In the A330 only ALT LAW is communicated by the ECAM, although ALT2B is latched. And i fear, that it is not en vogue to know too much details about systems and possible systems malfunctions, when flying an aircraft which performs almost all tasks on its own.

Indeed ... this may be inexplicable ..
The fact is that the BEA has investigated whether this behavior was explained (or induced)
He did not give a 100% satisfactory answer but gives in its conclusions and recommendations a few tracks

studi, perhaps one reason you can't explain what the pilots were doing is that we don't know what they were looking at. We don't know the actual scan pattern the man in the Right Hand seat was using, but some informed guesses can be made. I find the BEA's suggestion in re AoA indication to be of merit. One more tool in the tool box, to be used when necessary.

If you reach an understanding of what he was seeing on his display and what he was then looking at as various thing changed mode or failed to display, you can get an idea of what may or may not have been wrong in training and habit patterns, and on what basis he made various decisions in trying to handle the aircraft when the Robot conceded control to him.

ONe part of the mishap that concerns me is what looks like a non-trivial CRM breakdown between LH and RH seat. That can happen to any crew, in any type aircraft, regardless of what sort of robot assisted flying system is in place.

I suspect that Air France, Airbus, and perhaps significant sectors of the industry do not feel incentivized to make in depth systems knowledge a part and parcel of their pilot force. If nothing else, this crash ought to wake up anyone in management to how bloody important it is to know your aircraft inside and out, and in particular, anything and everything to do with flight controls, which in FBW aircraft means How The Robot Works. If the robot gets a share of flight control manipulation tasks, the pilot has to know what's going on, and when, without ambiguity, since at any moment the robot may admit that it hasn't got what it needs to fly and needs the pilot to take care of it.

Studi, I don't think it is possible for me tell you how strongly I disagree with what you said there. The requirement for Professional Pilots (see name of forums and community where this discussion is taking place) is a degree of care well beyond what you assert.

Hi studi....

You say Quote,

If I ever get into a protection, I will already have been so much overloaded that the deeper knowledge will not be readily available for use anyways.

That is precisely what happened to PF Bonin........the Stall Warn established the limit. And it happened within seconds of being handed the airplane.

Care to make a comment?

@Lyman:

Bonin only got the Stall Warn because he pulled up.

@studi:

I can't help but feel that something got lost in translation - could you repeat that in German?

(Leider ist mein Deutsch nicht so gut, aber hoeffentlich kann ich dich doch verstehen...)

B*****d....


Hi studi....

You say Quote,

If I ever get into a protection, I will already have been so much overloaded that the deeper knowledge will not be readily available for use anyways.

That is precisely what happened to PF Bonin........the Stall Warn established the limit. And it happened within seconds of being handed the airplane.

Care to make a comment?
*

Hi,

studi
I'm waiting the next .. and be sure it will happen .. it was already a close call some months after AF447 by another AF crew .. (I have the statistics working form me .. and if not enough .. Murphy will help :) )
You know ... never more Three Mills Island .. never more Tchernobyl .. never more Fukushima ... etc ....
The man likes a utopia .. the dream :rolleyes:

On this I agree to some extent - Chernobyl was caused by putting men who didn't understand nuclear power in charge of building and running the largest nuclear reactors in the world. The problem in aviation is the proliferation of management and executives who don't understand aviation.

Actually, the RBMK reactor had dangerous design flaws, that were completely unknown to the operators. The rods that were intended to stop nuclear reaction, were designed in such way, that they actually accelerated the reaction for a few seconds. That was the immediate reason for the explosion.

On some other stations equipped with RBMK reactors, the flaws were already encountered by the operators, but no modifications to the design of the reactor were made until the catastrophe.

Off-topic, but sounds weirdly familiar in this thread's context...

I know that - but the ultimate reason for the disaster had less to do with reactor design than politics. The operators were unaware of the design problem, but they were expected to abide by a protocol which said "do not have less than X control rods inserted at any time". The test they were running on the night of the disaster was supposed to have been done before reactors 3 and 4 ever came online (when in fact they'd been running for over a year). Viktor Bryukhanov, the plant director, had fudged the paperwork sent to Moscow to get them online early, but an inspection from Mosow was due in the following weeks. If the inspectors had found the paperwork missing, Bryukhanov would have likely been prosecuted.

So he ordered his deputy, Anatoly Dyatlov, to make sure those tests went ahead before the inspection come hell or high water. Dyatlov's only prior experience had been on nuclear submarine reactors - he had no concept of the power of the RBMK or the risks involved in operating it, so he ordered the test to go ahead over the objections of shift leader Alex Akimov (a young engineer who had studied the risks, but was not informed of the RBMK design problem, and who - sadly - was posthumously held responsible). He did so by threatening Akimov with dismissal (and, by extension, his family with destitution). Akimov gave in, removed the rods and the rest is history.

You say this is "weirdly familiar", but it really isn't. The Soviet Institute of Energy withheld information on the near-disaster at Ignalina, and the flaws in the RBMK reactor they highlighted. Airbus publicised the issues with the Thales AA pitot tubes and published a UAS procedure to deal with it, trusting the airlines to disseminate that information to the crews. They enforced a Service Bulletin to replace all Thales AA pitot tubes by a certain date, but unfortunately the accident happened anyway.

The "reason of Chernobyl" topic is the usual reason for holy wars on the respective forums, in the same way as "A vs. B" in aviation. There is much to be said to support my point of view, but let's not start it here :)

@ap08:

It is a greatly misunderstood disaster (in part because the report published in Vienna was a political whitewash), but that's not the issue. Airbus did not restrict information on the known issues with Thales pitot tubes.

One can see other similiarities between the two disasters. In both cases:
- operators did not fully understand the system (see Studi's post, pilots do not need to know how the system works to operate it safely...)
- operators made outrageous mistakes and violated the procedures (I agree that deep political reasons should be sought in both cases)
- but the system also responded inadequately (autotrim put the airplane in unrecoverable stall; control rods started the explosion)

Autotrim did not put the aircraft in unrecoverable stall, the PIC did by continuing to pull up. If the PIC (or PNF) held the stick forward for as little as 8 seconds the autotrim would have rolled the THS forward and *aided* the recovery. Not only that, but if the crew had levelled off at any point prior to the apogee of the zoom climb, the autotrim would have stopped the THS then and there.

People can argue back and forth over systems understanding till the cows come home, but the 800lb gorilla in the room will always be that the PIC responded to a transient UAS problem by pulling up, continuing to pull up through the resulting zoom climb and continuing to pull up through the stall - and that's before we get to the stunning lack of CRM during the sequence.

The point is that without autotrim, the stall would have been less severe, the pilot might have been able to recover or the nose might have dropped on its own.

Also, if I understand correctly, autotrim was NOT supposed to work this way. If the system knew the correct speed, trimming up would have been inhibited. In this case, speed was unknown and AOA/stall warning not taken into account by the system logic...

" you just fly your airbus as any other plane and you are fine".

It flies like any other plane only in that left/right, up/down SS movements generally translate to left/right bank and nose up/down pitching motions.

It is not flown like "any other plane"

Does "any other plane" have a no feel, no feedback SS?
Does "any other plane" allow left SS to fight with right SS and then allow the subsequent warning to be drowned in a cacaphony of sensory overload?
Does "any other plane" remove it's pilot from awareness of its trimmed condition?
Does "any other plane" trim for gee or flight path vs speed?
Does "any other plane" remove power awareness from its pilot with its non moving power levers?
Is any other plane" flown differently depending upon its flight control computer abilities? When I say "flown" I speak of the physical, tactile motor skills required to control an aeroplane.
How many other planes deny their pilots the ability to slip into a crosswind on landing?

I'm on record as stating that any pilot who states that FBW Airbii are flown like "any other plane" simply doesn't know how other planes are flown, and I'll stick by that statement.

I fly it, and now that I understand it's idiosyncrasies, I feel safe amd comfortable in it. But that doesn't mean that it flies like "any other plane"

The Cirrus SR series, F-16, Dassault Rafale and F-22.

The procedure in such a situation is to hold the sidestick override switch down, which will lock out the opposite side.

The FBW Airbus pilot can see the trim condition by looking down at the inboard trim wheel, just like any other airliner.

Plenty - it's been standard practice in fighters for decades.

You can see the power settings by looking at the "donuts" on the centre MFD - you know this!

(And opinion aside, moving thrust levers did not help the crew of Turkish 1951 or SAS 751)

Yes - the B777 and B787 for starters.

Please elaborate!

(and in the meantime, a couple of threads on "slipping" the A320)

http://www.pprune.org/tech-log/40377...lip-320-a.html

http://www.pprune.org/tech-log/15907...technique.html

http://www.pprune.org/tech-log/36188...-tecnique.html