As I said, I don't think the crack - if indeed there is one - falls between the designer/maufacturer and pilots (after all, pilots were involved in the design of the Airbus FBW systems). I think the disconnect is between airline management and the pilots that work for them, because the breed of management that started to proliferate towards the end of the '80s were pure MBA material, and the old-school management who saw aviation as a specialty all began to retire.

John T: good points!

As seems to have happened pre AF447, and probably still happens, sim "training," or more importantly sim sessions with "x" on the line, may not be designed nor used to do a "training" function but instead confirm adherence to standard policies and practices, or to highlight subtle changes or revisions to same. "Gents, as you know, there is a new SOP on CAT III approaches. You've read the bulletin, let's see that you know how to put it into place during this sim session!" (And other tasks, of course ... )

There is value for any organization in doing that. There is solid argument to be made against conducting training on revenue flights, since the pax didn't sign up for other than "get me from point a to point b."
Respectfully, I contend that any Airline company that does NOT invest in its human capital -- pilots-- as I had previoulsy suggested, in which actual Training, refresher, and currency are deliberate and well defined aims of their precious and expensive simulator time, are wasting time that they pay for. Not seeing the forest for the trees, perhaps.
Indeed. :ok:

I suggest that it is in the self interest of any airline to keep its crews sharp, and deeply systems knowledgeable. The better you know your systems, inside and out, the better you can do your job and complete your mission.

Adopting that attitude reduces their (company) risks and increases their (company) success rates of "malfunction/emergency sorted" when things do go wrong.

I do not believe that I am the only person who holds that position.

Problem is, I don't run an airline and thus am in no position to make an impact. That a clear and careful implementation of this philosophy, as you describe it, has to go hand in hand with how to fund it presents any airline management -- or leadership if there is any -- with a critical problem to solve.

Well, that's why those folks in suits are paid.

I have a few :mad: ed thoughts on what I think of MBA's who aren't pilots running airlines ... but that does not belong in this thread.

AF447 has to me shown that Air France (and perhaps a good portion of the industry?) does not make the points I allude to above a priority. (I may be overstating this, or may simply be wrong.) Maybe the idea I have would be that they have a system in place that doesn't achieve that end even if that were its intention.

Result: a crew unable to fly and CRM its way through a malfunction, with fatal results. From the CVR excerpts released, I get the impression of a crew that were trying and trying to get the situation under control, but they were trying the wrong things to resolve their malfunction. (And one could argue that the "how" of their trying, as a crew, was not what we have come to expect).

An analogy to some of my own experience would be something like this: in a helicopter while flying above the high seas, experiencing a loss of tail rotor control during IMC conditions, but treating it as loss of tail rotor thrust or an AFCS malfunction. Could have ended in tears, were I and my H2P to respond to such a malfunction, as a crew, incorrectly.

BOAC, re post #284, "What I am trying to establish is whether the UAS QRH action PROMULGATED BY AF at the time of 447 was as described."

If I may leap in on your question to Dozy, the QRH drill and checklist did not reflect the EASA AD Procedure, which is:

The QRH drill has essentially remained the same from November 2002:

IMMEDIATE ACTIONS
- AP/FD........................................OFF
- A/THR........................................OFF
- FLAPS........................................MAINTAIN CURRENT CONFIG
- SPEEDBRAKES............................CHECK RETRACTED
- L/G............................................UP WHEN AIRBORNE

IMMEDIATE PITCH ATTITUDE AND THRUST GUIDANCE
If the failure occurs before thrust reduction:
- THRUST LEVER...........................TOGA
- PITCH ATTITUDE.........................12.5°

If the failure occurs after thrust reduction:
- THRUST LEVER............................CLB
- PITCH ATTITUDE below FL 100.......10°
- PITCH ATTITUDE above FL 100.......5°

RESPECT THE STALL WARNING

WHEN FLIGHT PATH STABILIZED
- PROBE WINDOW HEAT...................ON
- ATTITUDE/THRUST........................ADJUST

The AD was to be incorporated into the applicable AFM 15 days after the effective date of the AD, (Jan 05, 2011).

Of interest is a note regarding the AOM, "the Aircraft Operating Manual is not intended to provide basic piloting skills, basic jet aircraft piloting techniques, or information that is considered as basic airmanship for trained flight crews familiar with that type of aircraft and with its general handling characteristics."

J.T., Re post #303, "Perhaps we should have a beer or ten to debate the pros and cons ?"
Re training vice checking, and jeopardy vice non-jeopardy, you should have seen the debates when we introduced LOFT!

The politics of failure and the underlying struggle for who controls who...regulator, airline, union, individual pilots - what a time.

Agree with all your comments on the sim. That said, I think there is reason to accept that "extrapolated" sim behaviour at and in the stall may not be that far from the airplane even though the regime is unflown, untested and has no data from such events. Because even a Level D isn't certified for here-be-dragons territory, it could only be used informally during any training. The primary difficulty has been the methods by which the AF447 circumstances are repeated, because the loss of pitots was not a separately-programmed scenario so the results had to be simulated using other methods and that can produce system behaviours that may not obtain in a loss of pitot(s) data.

The rules haven't changed, so an airplane that doesn't meet the stall characteristics requirements without a stick pusher will still need one. What you have written applies to FBW airplanes. The FBW system permits modification of the natural characteristics of the airplane so that it meets the requirements. The Final Report describes the type certification in section 1.18.1.

While the airplane flight control system is designed so that, in normal law, the pilot or the A/P cannot command an AoA greater than alphamax, and therefore cannot stall the airplane in the 'usual' way, the system apparently can provide a powerful nose-down pitch very similar to a stick pusher if alphamax is exceeded in any other way.

Hazelnut39
Indeed .. in normal law
What about the other laws (no more AOA protection) isn'it the return to a "normal plane" ? and so need stick pusher ? ...

@HN39

Agreed, but what Lyman was referring to was a physical stick push that actually moved the column/sidestick.

@jcj:

I'm pretty sure that the A330 meets the stall requirements such that a stick pusher is unnecessary. As I said, stick pushers were a technical solution to the T-tail "deep stall" issue, and while they were installed in some non-T-tail designs some decades ago, that was to satisfy a UK CAA mandate.

PJ2: as I understand your post, the procedure you cite in the blue box is based on lessons learned from AF447, while the other procedural steps were in place at the time (per your observation "... essentially remained the same from November 2002") of the mishap. What I don't have clear is which symptoms were considered sufficient trigger to enter that series of procedures. (Airspeed unreliable, or A/P kicked off due to Airspeed Unrealiable?)

To ensure that I am speaking the same language:

Does "immediate actions" mean "memory items" or is that "break out the checklist/procedure and do these first" in a challenge/response style of ordered operations by the crew?

IMMEDIATE ACTIONS
- AP/FD........................................OFF
- A/THR........................................OFF
- FLAPS........................................MAINTAIN CURRENT CONFIG
- SPEEDBRAKES............................CHECK RETRACTED
- L/G............................................UP WHEN AIRBORNE

IMMEDIATE PITCH ATTITUDE AND THRUST GUIDANCE
If the failure occurs before thrust reduction:
- THRUST LEVER...........................TOGA
- PITCH ATTITUDE.........................12.5°

If the failure occurs after thrust reduction:
- THRUST LEVER............................CLB
- PITCH ATTITUDE below FL 100.......10°
- PITCH ATTITUDE above FL 100.......5°

RESPECT THE STALL WARNING

WHEN FLIGHT PATH STABILIZED
- PROBE WINDOW HEAT...................ON
- ATTITUDE/THRUST........................ADJUST

I presume that "adjust" in that last means "adjust to desired speed and power combination for continued flight."

DW
The "approach to stall" requirements .. certainly (regulators requirement fulfilled .. or this plane can't fly commercially)
"The stall" requirements (if any) .. we don't know ... the A330 in stall situation was not tested ... AFAIK
BTW AF447 can't be considered like a valid test ..

Where do you have this information from ?

I would be very very surprised if the A330 wasn't stall tested. I know that other models have been (see eg. about the A380 testing - see the comments on stall in the full description on youtube) and as far as I know it is required for certification.

The A330 (and in fact any airliner) will not have been tested to the AOAs that 447 went to (this may be what you are confused on?). Nor will it ever. As I think someone else already posted recently, no one would be suicidal enough to try the 447 stall in the real a/c (and therefore there is no real data for recovery profile in the Sim either - it's all extrapolated).

Conditions other than normal law are associated with system failures. Section 1.18.1 of the final report describes how these are addressed in the certification requirements.

infrequentflyer789
I see at 18 sec (and more) in the video people with a parachute on their back .. :)
So .. if this is possible (apparently) to evacuate the A380 in case of danger .. I see nothing suicidal to test stall like the of AF447 (the most risk is to end with a smoking hole in the ground)

My original challenge to Doze was re: the Shaker. The recommendations of BEA re: STALLWARN gave me the impetus to explore the elements in the region of Approach to STALL.

The STALL is interesting, but to me, I still believe the LOC was made possible by actions and mechanical insufficencies prior to the second STALL WARN. In Alternate LAW, why should the A330 be allowed to avoid installation of the shaker? Can we not see that an additional and different sensate device would have saved the day here?

What pisses me off is the sacred cow nonsense I see in the Certification side. Is it impossible to equip a SS with an annoying vibration that would startle the pilot into focusing on the impending death of his passengers and loss of his aircraft?

Without protections the 330 is no different in controls theory than non airbus vehicles. Besides, the Bus seems to know well enough not to remain in NORMAL LAW if it is in danger of STALLING. Very convenient as a sales tool. Impossible to STALL? Read the gd fine print, eh?

Well no, as shakers have been ignored several times in aircraft equipped with them, sometimes fatally.

Look, Sir, it is getting tiresome that you expect to make a glib comment about a failure of some piece of equipment and expect others to accept your pronouncement as 'end of discussion'.

Nothing works perfectly, consider the STALLWARN in this accident, ok? Sometimes engines quit, shall we give up flying?

Why don't you take an objective guess at whether the Shaker may have helped here?

Lonewolf_50;, Re, The term "Immediate actions" refers to the memory items only.

As we know, this drill has changed over the years. Until sometime in 2008, the drill began with "Immediate Actions", with no qualifying condition. There's info on this on the web or in the IR/Final Rpt as to when it changed but I don't have the time at the moment to source it.

Sometime in 2008, the qualifying condition, "If the safe conduct of the flight is impacted:" was added to the same drill items.

The way the drill was then written effectively split the drill into two parts: "Impacted/Not impacted". If "impacted", one does the boxed items. If not, the next "dot point" is "To level off for troubleshooting", etc. I don't think this was ever really clear prior to AF447. This is partly because within the memorized items under the qualifying phrase, was a qualifying condition that catered to low altitude failures but would not be done if the safety of the flight wasn't impacted: "When at, or above MSA or Circuit Altitude: Level Off for Troubleshooting". So the guidance was confusing, in my opinion and I thought the immediate actions should be qualified by flight-phase, not the very subjective, "If the safe conduct...", etc.

So there is no real, obvious guidance for cruise ADR or UAS/pitot failures because the drill was originally conceived after the Birgenair and Aeroperu accidents.

Trouble is, UAS events on the A330/A340 in cruise first occurred prior to April, 1996 and many listed in the BEA reports had occurred by 2008 yet the drill remained unchanged. I think appropriate training would sort this drill and checklist out such that it could be done correctly but sorting it out for oneself leaves the drill open, as we see, for interpretation as to what to do and when.

The addition of this qualifying condition placed the notion of "Immediate Actions" in limbo...What was one to do first?

At takeoff, it is clear - the safe conduct of the flight is impacted and one does the memorized items.

In stable cruise, even with mild turbulence, what does one do? "Above FL100, set pitch to 5deg, (about 2.5deg higher than normal cruise) and get out the books, but was this ever trained?

I'm won't re-argue the case but in response to your and BOAC's comments, aside from the addition of the qualifying condition, the drill hasn't changed, and still (to my knowledge) has not incorporated the AD admonition to not engage the AP or FDs, and not to follow the FDs until at least two airspeeds are considered valid.

Lots to chew on yet.

Here's the 2008 QRH UAS drill:


http://www.smugmug.com/photos/i-sm2J...sm2Jtgb-X3.jpg

Lyman, Dozy is against anything that "might have helped."

I can only be relieved that he's not in charge of any airline.

:bored:

Hi PJ2,

Relative to the QRH drill, I have some questions. As in the case of AF447, you are cruising at 350, M .82, 2.5º pitch and an N1 setting of 95% (as referenced on page 183 BEA's final report for level flight). Due to expected turbulence, a reduction in M is dialed in to .80 resulting in a reduction of N1 to 84% by AT. Suddenly the AP disengages, the AT disengages and AT lock occurs. At this point, according to the QRH drill, the pitch should be set at 5º and the thrust levers must be moved to climb detent at which point one needs to know from the tables to reset pitch and power based on the aircraft wight and altitude. My first questions would be aren't the thrust levers already in the climb detent? Does the thrust lock latch a 84% N1? if so, does this give much time to figure the right pitch and power setting before a real stall warning occurs?

But when I see the tables, the reset for pitch is 3.5º and 90% N1 for a M of .82 based on altitude and weight. This seems a little low on N1% with a 3.5º pitch setting based on what is referenced in the BEA report for level flight at the altitude being experienced. I recognize at the upper end of the power settings at 350 there isn't much power left to be had, but the differences seem confusing to me more than anything, perhaps I am missing some finer points. Any thoughts? Is my logic off?

P.S. The source of pitch and power table I used came from the BEA final report, Page 104. Additionally, the BEA Interim Report #3 gave the following N1% power settings from each engine as follows:
At M .82 - True N1% 100.4/99.8
At M .80 - True N1% 98/98

No, I'm against blanket assertions that something *would* have helped when there is no evidence to support said assertions.

Adding complex mechanical devices would add more potential points of failure to the flight control system for no proven gain.

Yet again I can't help but suspect a slight bias towards "the Boeing Way".

Organfreak Lyman, Dozy is against anything that "might have helped."

I can only be relieved that he's not in charge of any airline.

Unfortunately, the rhetoric that is being parroted is that of people who build and certificate the AIRBUS.

The shaker is experienced thousands of times a day in training, and typing.

"No evidence there would be any gain?" What a wild statement.

I don't know about thousands - maybe hundreds - and the same could be said of the Airbus auditory/visual warnings. Unsurprisingly, in training situations, pilots respond well to both.

No, it's a true statement unless you can provide evidence to the contrary - a "wild" statement is insisting something would have helped when there is no evidence to support it (and plenty of evidence that overwhelmed crews ignore the shaker just as easily as any auditory or visual warning).

Hello TD;

Re post #330:
Yes, you're correct - the TLs would be in the CLB detent already. The drill merely ensures that this is so, just in case they have been moved out of the detent for any reason.

Yes, again correct - the Thrust Lock would synchronize with the aircraft parameter at the time of disconnect. Eighty-four percent is sufficient to maintain stable flight for the time needed to bring out the QRH, (which should always be right at hand). The page hasn't changed in quite some time, (2.21 in Nav Abnormals). In fact a five-degree pitch attitude results in a gentle climb of around 800 to 1200fpm, (again, I was wrong on the original statements) - for an experienced crew, this would not present a problem as "close to nothing" has changed and one can return to level stable flight with tiny movements on the stick and observing the altitude and altitude trend, (rate of change). Digital vice analogue adds a cognitive step because the notion of "amount-to or -away from" is missing - no big deal.
No, your logic is sound - it takes a significant change in pitch or power to place the aircraft at risk of overspeed or stall during this drill. The settings given are very close to what the aircraft would require but they're not exact. Going from recollection of the FCTM here...I think an answer lies in the 10T weight category intervals. The FCTM does state that the settings are approximate and that fine-tuning involves monitoring altitude using either indicated or GPS altitude:

LEVEL OFF AND STABILIZATION (IF REQUIRED)
The table gives the proper pitch and thrust values to stabilize level flight according to weight and altitude.

If the altitude information is unreliable, the FPV and V/S are also affected. In this case, the GPS altitude, if available, is the only means to confirm whether the aircraft is maintaining level flight.

When the altitude information is reliable, use the FPV.

 If the FPV is reliable, or if the GPS altitude is available:

‐ Maintain level flight (FPV on the horizon or constant GPS altitude)
‐ Adjust thrust according to the table
‐ Observe the resulting pitch attitude, and compare it with the recommended pitch target in the table:

• If the pitch necessary to maintain level flight is above the table's pitch target, the aircraft is slow, then increase thrust,
• If the pitch necessary to maintain level flight is below the table's pitch target, the aircraft is fast, then decrease thrust.

When the pitch required to maintain level flight gets close to the table's pitch target, re-adjust thrust according to table's thrust target.

This technique permits to stabilize the speed quickly while maintaining level flight.

 If the FPV is not reliable and the GPS altitude is not available (no means to ensure level flight):

Adjust pitch and thrust according to table values, and wait for speed stabilization. Expect a significant time to stabilize and important altitude variations during the stabilization.


Hope this helps.

Stall did occur at ~180 -190 kts IAS. That is around M0,55.
That is a huuuuge way from M0,8 or 0,82 from a kinetic energy perspective.
So even at 84% you're lightyears awy from a stall.
It takes a significant climb to shed that energy. Thrust variation (short of a shut-down) has much less immediate impact on kinetic energy than a climb. If simply keeping altitude you have ample time to sort things out.

Edit: Stall warning would come on earler especially in Alt Law but with a constant altitude it should come up somewhere around 230 kts (M0,65 - 0,7)

PJ, thank you very much, and also thanks to TD and henra for some amplifying thoughts.

I better understand what was, what is, and an idea for "what ought to be."

The industry as a whole, who fly AB 330's, will hopefully come to a consensus on a "best practice" regarding this malfunction family. What I gather from your observations is that perhaps the consensus has not yet been achieved.

Hi DozyWannabe,

Please see Page 105 of the final report:

"1.16.8.3 Response to aural warnings
Numerous studies have been conducted on insensitivity to aural warnings and they showed that the aggressive nature, rarity and unreliability of these warnings may lead operators to ignore these signals [1, 2]. In particular, in the event of a heavy workload, insensitivity to aural warnings may be caused by a conflict between these warnings and the cognitive tasks in progress. The ability to turn one’s attention to this information is very wasteful as this requires the use of cognitive resources already engaged on the current task. The performance of one of these tasks (solving the problem or taking the warning into account) or of both would be affected [3].
In addition, studies on the visual-auditory conflict show a natural tendency to favour visual to auditory perception when information that is contradictory and conflicting, or seen as such, of both senses is presented [4, 5, and 6]. Piloting, calling heavily on visual activity, could lead pilots to a type of auditory insensitivity to the appearance of aural warnings that are rare and in contradiction with cockpit information.
A recent study in electrophysiology on a piloting task seems to confirm that the appearance of such visual-auditory conflicts in a heavy workload situation translates into an attention selectivity mechanism that favours visual information and leads to disregarding critical aural warnings [7]."

Having the stall warning alert only via the auditory channel is not a good design during high work load situations.
If you were busy arguing with Lyman in a noisy party and I wanted to get your attention, the most effective way would be by touch (on your arm or shoulder say). It's a completely separate channel with no competition from any other warning - as is the stick shaker.

Hi PJ2,

Thanks for the response to my questions, much appreciated.

From what I've read I don't believe that Dozy was saying a Stick Shaker wouldn't benefit. Simply that you can't look at it from your armchair and say "that's the solution". You need to look at the whole system and work out the pros and cons. If I read the report correctly, in this tragic situation the pilot ignored the aural warning, ignored the instruments and even ignored the other pilot in the cockpit. He had become fixated completely and if he was ignoring all of those I would guess, and it is a guess, he had a mental picture where he no longer believed what the systems were telling him, if that is the case then there is a real possibility he would have ignored the stick shaker as a false indication too.

As for the crowded noisey room scenario, have you ever tried to break up a fight just by tapping someone on the shoulder?

I will also question why some people on this thread are obsessed with a stick shaker. This aircraft had a system that could have safely restored level flight had it been re-engaged. So, instead of putting a stick shaker in, why not allow the system to take control again if it is convinced that the aircraft is in danger of exiting the envelope?

The system is designed to respect principles that have been held sacred for several decades. The safety analysis is extremely stringent and forces decisions into the design that when analysed against a very specific set of conditions doesn't seem to make sense. If you step back and look at the whole picture, with all the analysis at your disposal, it makes more sense. Of course nothing is perfect, and it takes such events as this to expose weaknesses in the design decisions. However in these cases it is very rare for the fix to be simple. The system has to be re-analysed.

I have read a lot of comments on these threads, many seem to be from people who, from what they've written, appear to have never been involved in designing, clearing and certifying a complex avionic system. But even in their position of ignorance they are quite happy to throw rocks at the Airbus design team.

In my view of this tragic event, the aircraft let down the pilot by suddenly exposing him to a set of conditions that caused him to lose situational awareness and expecting him to run with it. The pilot let down the aircraft by not carrying out the basic tasks required to return the aircraft to safe steady flight. The reasons this happened are no doubt deep and wide going way beyond just the pilot and the aircraft, the lessons will be many and varied and hopefully will lead to changes that improve safety all round.

What scares me is for how little a span of time the airspeed was unreliable and how quickly this event went from routine to catastrophic. What saddens me is that perhaps in 99 time out of 100 this pilot would have correctly interpreted the situation and it would have become a none event.:sad:

Which is why they've recommended bumping up the visual component, at a guess... :)

Heh.

I'm well aware of the theory that the tactile channel - as separate and distinct from audio-visual - can get attention when audio-visual can't, and it makes sense in terms of pure logic and psychology. This doesn't alter the fact that in times of high stress workload in the flight deck, shakers have been ignored multiple times. The only way to know for certain would be to blind-test the two methods, but I don't see that happening any time soon.

Ladies and gentlemen, dear PPRuNers, honourable participants in our discussion, I would like to take an opportunity to kindly ask you to get acquainted with The Book:

http://www.pooleys.co.za/images/Hand...big%20jets.jpg

It will do you a lot of good and also might help reduce the PPRuNe bandwidth wastage.

If you have third edition on your bookshelf, you might find the following chapters useful:

Pg 109, Stalling, which shows that pilots' wildly divergent views on the subject of stalling have preceded PPRuNe by decades. Also displays DP Davies fantastic ability of insight, in matters not just technical but also in human nature.
Pg 128, Stick shakers.
Pg 130, Stick pushers.

Stickshakers are not supposed to be universal stall warning devices! They are fitted to aeroplanes that have no pronounced natural pre-stall buffet or margin between buffet and stall is to thin for buffet to be used as effective deterrent. Why A330 doesn't have it installed? Two possible clues in the final report.

So 330 both shakes out of her own accord and did receive some special dispensation in certification process by virtue of her envelope protection.

Now about stick pusher: it is not stall recovery device, it is stall prevention device! It has to be fitted to aeroplanes with unacceptable stall characteristics. e.g. pitch-up at stall or prone to entering the deep stall. As anything else related to aeroplanes, while solving some problems, they bring in some problems of their own (HTBJ, pg 131) so claiming they are stall recovery devices and suggesting they should be installed on any aeroplane to deal with the problem of distracted crew would be sparkling display of aeronautical ignorance.

Since for nine years I have been earning my daily bread flying the aeroplanes equipped with both stick shakers and stick pushers, implying if I ignored the stall warnings and somehow pulled them into stall they would bite without warning and once stalled would stand good chance to be unrecoverable, one might perchance find understandable that I view the propositions to indiscriminately train each & every airline pilot in full stall recovery procedure with considerable amusement.

Yea... it was measly 13 degrees. Far too high for cruise anyway but compatible with the ardent desire to climb and having no idea aeroplane has insufficient performance to fulfill her masters wish - yet she attempted to obey into her eventual destruction.

Emphasis is mine but this is one of the most precise hitting the nail on the head that can be found on the PPRuNe.

Maybe it's just me, but when someone counters to spectacularly wrong theory with some solid facts disproving it, least I expect it would mean the end of that particular direction of discussion.

Sorry Dozy, I missed the part where this was discussed. Could you provide the reference that states latching of ALT2 permanently disables the autopilot? I need to get my head around this.

Lets not get too deep into the helpfulness of stick shakers.

There is some unique combinations at play in this accident

For starters its at high altitude and preceded by speed drop outs.

My feeling is that if only the stick shaker had annunciated, the crew would likely have addressed that bit somewhat better. But when multiple annunciations are alerted the tendancy of some might be to go to lowest common denominator in their memory bank ( and forget about the stick shaker or auto-something disconect)

I've seen this happen before

Indeed .. and the solution was simple ... and repeated by many here "Pitch and power"
What is needed for a auto control flight like in the Airbus to automatically go in a mode (one more law :) ) "pitch and power" in case of UAS ?
Can this be implemented ... for avoid the "1%" pilot failure to set the famous "pitch and power" ....

Stick shakers help because you immediately release the back pressure or decrease bank angle to stop it. I had it happen at night in a 727 over the water held high with speed brakes and a 30 degree bank circling to keep the pattern tight to lose altitude landing at POP in the Dominican Republic. The speed was fine but the bank and speed brakes set it off.

You instinctively fix the situation not ever experiencing it in an airplane, only the sim during training. It would definately override any oral warning.

jceant, #344

This is really the common thread of what i've been thinking in terms
of a system that should be designed to degrade more gracefully at
the edges. When all the standard "protections" fail, there should be
absolute limits in command inputs that cannot be exceeded to ensure the
safety of the aircraft.

With so much technology built into these machines, it amazes me that there
are so many corners where the pilot can get himself into serious trouble,
for whatever reason. The acceptable limits under all conditions of flight
are known, so there's no technical reason why such last line of defense
protections shouldn't be built into the system design.

Perhaps they never thought it could never happen, but it does...

Clandestino;It has been posted a number of times, but it was takata in AF447 - Thread No.5. I think A33Zab has also had something to say on the latching of ALT2B.

EDIT :: If you are looking for an AI definitive reference, then I have not been able to find it in FCOM or Instructor Training Manuals, so assume that it will be hidden somewhere in the AMM.

Maybe a bit semantic, but it is primarily a stall identification device. I fail to see the the distinction you make between recovery and prevention. On some airplane it is set to fire after the point of maximum lift (see fig. 5.25 in HTBJ) to provide the stall identification prescribed ("a nose-down pitch that cannot be readily arrested") with minimal loss of performance.

From HTBJ page 111 (my bold):
And on page 131:

If you want a stick pusher equivalent on airbus, just program it into
the software..

The more relevant question might be: Why hasn't this been done
already ?...

Thanks mm43, takata discussed the reasons for latching the alt law but I am looking for the confirmation of notion that has made circles around this thread lately: namely that latching of ALT2B disables the autopilot until ground reset. What logically follows is that since other crews normally used autopilot while remaining in ALT mode after regaining the normal airspeed display, they never got as far as ALT2B and AF447 was pretty unique. However, pretty old FCOM I have shows ADR disagreement doesn't result in AP being lost, so I'm looking for further reference.

Maybe semantic but it looks to me like a scope issue.

I fail to see how that follows from the foregoing considerations, and it conflicts with what follows immediately next.

Clandestino...

"Maybe it's just me, but when someone counters to spectacularly wrong theory with some solid facts disproving it, least I expect it would mean the end of that particular direction of discussion."

No solid facts were in evidence, merely a retort that a shaker had on occasion been ignored. Further, even BEA had trouble identifying Buffet and had to with accelerometer data from DFDR.

The dispensation you reference is what is galling. Do you honestly think the shaker would have been without possible benefit on 447?

The 330 in this case showed signs of Buffet only two seconds after the SW fired.

You are without doubt an expert, and I admire and respect your knowledge. Your personality is of no concern to me, I actually enjoy your posturing and preening. I expect that in a good pilot.

You need to read what I said again, which was not suggesting that a
faulty source continues to be used, but that tracking the two sources
on a continuous basis provides a more accurate view of the overall
situation, a potentially earlier warning of developing problems and
fallback in the event of a single source failure. Perhaps i'm not very
good at decribing what i'm trying to get across ?.

If you have an ivsi on the panel, then you are already using baro /
inertial mixing, though not in a particularly complex way.

Separate functions, yes, but cross check between the two to improve
validity and consistency of data.

That's high level detail and not really my area. No argument
about the last line though :-)...