Hi DozyWannabe,
If you read the cvr transcript, both crew knew they were close to the stall.
Air Florida Flight 90 - Wikipedia, the free encyclopedia
“16:00:45 CAM-1 Forward, forward, easy. We only want five hundred.
16:00:48 CAM-1 Come on forward....forward, just barely climb.
16:00:59 CAM-1 Stalling, we're falling!
16:01:00 CAM-2 This is it. We're going down, Larry....
16:01:01 CAM-1 I know it!”

Birgenair Flight 301 - Wikipedia, the free encyclopedia
“The co-pilot and relief pilot both seemed to recognize the approaching stall and tried to tell the captain, but did not intervene directly, possibly out of deference to the captain's age and experience.”

That's a CRM problem. The other pilots recognised the approach to the stall.

Reliability and performance are not the same thing. No one expects any system to actually cause a problem. The key is how is the performance, when there is a problem.

Two different things. When in a bad deal, what is the performance expectation?

To say that one system is not compromised in its design consideration because a separate system has been present in a fatal, is frankly preposterous. In the vast majority of cases, controls are not an issue. Can controls become an issue? Yes, I would say.

"Climb, then" "But I have been pulling for some time". Right. For some length of time, unknown to the other two crew, PF was inputting up elevators.

Can they have their time back? Request a do-over? What could have been the solution had the PF's yoke been in his lap, visible immediately at any given time?

Rest

It's been a while since I read about Palm 90, but I'm pretty sure that F/O Pettit knew they were stalling from reading the ASI, and that the Birgenair F/Os were using a combination of the ASI (which was working on the F/O's side) and ADI (the other F/O in the jumpseat calls "ADI!" out several times during the sequence). Neither made any reference to yoke position.

I know they were aware they were close to the stall. The point I was making was that even with the yoke telling them that their respective PFs were pulling into the stall, they either did not see it, or saw it and didn't do anything about it.

As non-flying pilots, their duty is to notice attitude, not the controls that produced it. "Yoke,Yoke"....... That makes them an instructor, and is not in the quiver of a highly charged cockpit. Even in emergency, habit and tradition.

"not acting" and "acting" are on two sides of a VERY thick coin.

Case in point: 447 PNF, apparently aware, and 'irritated' does not say, "Use your Stick, Push Nose down." He points to VS, and, due protocol, waits for the correction. None is forthcoming, yet he persists, unto STALL, apparently. That part of Training may need a looksee.

Lyman, I think you are stretching a bit too far there.

Look at the nose pitch behavior: it appears that the crew on the flight deck did not correct the nose attitude, nor maintain assigned altitude, within normal parameters during the early course of the malfunction, before it became something worse. It began as a comparatively mundane malfunction of a system known in the past to have some vulnerabilities. It does not appear that the malfunction they were faced with was novel, nor a "first time in class" malfunction.

Given that with similar malfunctions, most crews in 30+ similar events, who dealt with attitude and altitude excursions of varying magnitude ...

1. got their planes sorted out
2. remained flying
3. remained unstalled
4. prevented a malfunction from becoming an emergency.

then this crew's inability to do so stands out.

I do not wish to disrespect the dead, but I think that, with some of the info available to us via released CVR extracts, the crew has at least some responsibility, even if a BEA conclusion is reached that training, command decisions by the captain, documentation, systemic problems at Air France, probe replacement schedules, et cetera, are major contributing factors.

That "zoom climb" is to me doubly inexplicable, given that the crew were very aware of their altitude limitations for that sector of the route. Their discussion was recorded, to the effect that actual versus forecast temps at altitude did not permit a climb to a higher altitude that they had planned for (or left as an option) before takeoff.

The Airspeed was unreliable, but Altitude was not.

I cannot think of a professional airline pilot who considers a 1000 or 2000 foot altitude deviation (no less a 3000 foot deviation) as something other than very serious, and in need of immediate correction unless something else very untoward is also occurring at the same time.

Respectfully, no.

In multiplace aircraft, when the other guy isn't doing great at flying, you do indeed resort to "instructor" type techniques and behaviors in order to get him back into the ball game. In some cases, you have to take the controls, but more often it is simpler to talk him back into the game.

"Lower your nose ...
Roll left ...
Ball's out to right, step on it ...
Add power...
We are sinking -- POWER POWER POWER! ...
The Gear's UP! Wave OFF! (Go Around!)"

All of these are solid actions / calls that in some situations must be initiatied by a good co-pilot (PNF) that are also "instructor" actions/calls.

The two roles overlap; they are not mutually exclusive.

I make no moral judgment as to these two accidents, and you are correct. BUT, going from the evidence, can you say some political reluctance was not in play? You quote proper procedures, certainly. In these tragic accidents, I believe your "best practice" was not folowed, as it should have been. Do you see these as "best practice," then? It is occasionally extremely difficult to act without "equality" in the cockpit, that is my point.

As to 447, "Responsible" remains to be decided, and since it is in Court, some percentage value will be apportioned to whomever the principals are determined to be. If crew were briefed, officially, to ignore STALLWARN, where is their culpability? So that means that up til three seconds after a/p handoff, the pilots were totally on the hook. The first STALL,STALL is intriguing, and we do not know its genesis. PF did not "TOGA/15 degrees at this point, and did later, so to me, that indicates the first warnng was ignored. If PNF was concerned about something "What was that?" we do not know what it was.

fifty four seconds total of STALLWARN, and not a word......do you think that odd, Lonewolf?

[B]@Lonewolf...."That "zoom climb" is to me doubly inexplicable, given that the crew were very aware of their altitude limitations for that sector of the route. Their discussion was recorded, to the effect that actual versus forecast temps at altitude did not permit a climb to a higher altitude that they had planned for (or left as an option) before takeoff. /B]

Not inexplicable and the reason is this: The evidence. My take is that the PF was unaware of his climb, and his instruments were not helping him to decide the correct attitude, plus a concern for Overspeed. Another factor could be an uncommanded ascent. Make no mistake, his pull on the stick caused climb, but can we eliminate the a/c climbing on her own? In additiion to? The VS and the altitude gained are remarkable, and the initial inputs of PF were not commensurate with ROC, imho. How can the a/c sustain 1.68 g and the cockpit not remark? ROLL masking the VS accelerations? Columnar airmass, ~ +100?

Lyman,
B-but Lyman, that was supposed to have gone out with high-button shoes, or at least hula hoops, after numerous accidents were blamed on poor/no CRM. How could, in this day and age, cultural customs trump professional procedure???

And, if there was such a problem, it seems to point to: they would rather die (and kill everyone else) than speak out-of-turn! It's hard to fathom. I'm wondering something that can never be answered: Were they both so panicked that they froze and forgot everything? (Deer, headlights.) It sure looks that way.

Howdy Organfreak.

Given the evidence, that the yoke was certain salvation, and nothing of consequence happened, you are correct in assuming some sort of hard to explain vector of behaviour. I count three times that I was very close to dying, once in a hang glider.

I did everything I could do, in all three situations, and each time I experienced a very calm feeling; in looking back, I could not have done anything differently.

BTW. I'd like to recapture that feeling, and like most humans make it my way of life. But without the danger.....

Hi,

DW
The question can be also ... why Boeing keep the yoke even on their last aircraft conception ?
Why they don't choice the SS ? .. as it's less weight .. etc ...

Diesel8:
Well, my friend, you made me go back to my logbook only to find out that I have about 7760 fh on FBW aircraft and 7740 fh on non-FBW.

I guess I'm a lucky guy ;)

In part because it gave them a selling point by taking advantage of the backlash among some following the A320's entry into service. FBW as a concept was already proven, so they could focus on the backdrive software, which was probably more complex than anything Airbus could have attempted back in 1982.

That was all window-dressing really though - where Airbus had and still have an unassailable lead over Boeing is the ease of conversion between types - in which the design of the Airbus FBW flight deck played a crucial role. Because there is more commonality between the Airbus FBW types than could be remotely possible between Boeing's, it didn't make sense for Boeing to try to retrofit all their existing types with 777-based controls. In that sense, Boeing having been around for longer and being responsible for the most popular airliner in the world actually proved a disadvantage.

[EDIT : Just to try and clarify - the debate over Airbus, FBW and automation frequently misses the point in a lot of ways. The A320's automation is not a great deal more advanced than that of the 757 (Boeing's 3rd-gen narrowbody at the time).

The fact is that Airbus found themselves in an almost unique position in the early 1980s. Because their only existing types catered for a specific market, and because that market was near-saturated, it meant that they could expand without encumbering themselves with A300 commonality.

One of the great misunderstandings about Airbus's much-publicised cost savings to airlines with the new FBW types was how the savings were achieved. It wasn't really to do with greater automation, and it definitely wasn't an attempt to de-skill pilots - it was to do with ease of conversion. By designing the flight deck of their narrowbody and new widebody types to have unprecedented commonality with each other, theoretically the cost to airlines of conversion training could be drastically reduced. Boeing had made inroads on this theory by designing the 757 and 767's flight decks to be very close functionally, but Boeing still had the legacy types (727, 737, 747) to contend with. Airbus had a relatively clean run at doing it across their entire range, with the exception of the A300/310 (which had nowhere near the market penetration of the older Boeing types).

As such, the fundamental design brief was to put together a flight deck that could be installed in anything from a twin-engined puddle-jumper to a four-engined widebody long-hauler and as near as possible have the only difference to the pilot be the reference numbers. Obviously if you're feeding back the control surface resistance to the stick that's not possible - so with the help of some of the best pilots in Europe they set about making a design which would feed back all the information available in a more traditional setup, but would fulfil the design brief. Some compromises had to be made, but the position Airbus enjoys now is a testament to the soundness of the design they came up with. ]

I think the Bus markets well, but let us not leave out the ease of transitioning to flight in the first place, the entry level. How much of this "consumer friendly" sales Pitch has to do with ease of operation? That is the foundation for the chronic knock on the Bus cockpit, imho.

All of it, Lyman - that's why people like Gordon Corps were onboard.

Leaving aside the sidesticks for a moment, the rest of the design was probably one of the most ergonomically friendly in airliner history. One of the reasons I miss 411A is because he could wax lyrical on what a pig the early 707 cockpits were (and different for every customer initially!).

The fact that I could take off, follow the FD and land the sim with minimal instruction* - having not been in a real aircraft cockpit since I last climbed out of an AEF Chuckmunk 18 years ago - shows that it is pretty intuitive to the novice.

[* - Not to underplay the instruction I did get, which was concise, to the point and everything I needed to do it - I can't thank those involved enough. ]

Lyman re your post of not reacting to Stall Stall.
A lot of years ago I was involved in an investigation into a wheels up arrival.
The GPWS was triggering for 17 seconds. None of the 3 crew heard it. However it was as clear as a bell on the CVR area mike. 1st thing to go under stress is hearing so I guess the 447 didn't hear the stall warning.

jcjeant wrote:
I realize that your question is only rhetorical, but I have read recently (but can't remember where, dammit!) Boeing's statement of their FBW design philosophy. It made it clear that they insist upon realistic FC feedback.

The question then becomes, why do they insist on that - given that the realistic feedback is entirely artificial?

I had a pop at answering that question above. I believe it has more to do with corporate philosophy and the need to stake out different ground from their competitors than anything else - plus they weren't in a position to reap the benefit of radically overhauling their control methodology in the way that Airbus were.

@IcePack - there are visual stall warning cues too, including the bright red master caution light going off right in front of you and various cues on the PFD.

(And whatever else you may think of Airbus, they were on the money - going from a distant third in terms of western airliner sales to being able to duke it out with Boeing for the crown on an annual basis.)

Lyman (below) - spelling corrected.

Chipmonk....that is a Catholic deHavilland?

DW:
Come on, Dozy, you know the answer as well as anyone. Because it is entirely realistic. It does its job, it works.

As to your second point, I'm sure that's true.

DW:
Great Jumping Jehosaphat! That is so disingenuous I forgot to sneer. MD merged into Boeing. That pretty much took care of the rest of the competition. (Though Boeing titularly bought-out MD, Boeing's board and leaders suddenly became largely MD, and that's when they began dropping lots of balls. Run by money wonks instead of engineer wonks.)

But I digress.

Because American pilots are more conservative than their European collegues?

Until it doesn't (just because it hasn't gone wrong yet doesn't mean it won't). One of the things I used to find amusing about the A v B debate were the B people who swore up-and-down that Boeing's latest models weren't entirely computer-reliant, and the presence of the moving yokes proved it.

They'd overhauled the ailing MD by the early-mid '90s. In any case the point still stands, because the MD/Boeing merger would have produced a company that consistently outsold Airbus year-on-year if Airbus themselves hadn't done something about it.

[EDIT : I'm sure I've said this before, but as an aside, I wonder what would have happened if Boeing had elected to shrink the 757 rather than stretch the 737 for their (then) next-gen narrowbody. ]

Hi DozyWannabe,
http://www.davi.ws/avionics/TheAvion...ook_Cap_11.pdf
"11.3 Design Philosophy.

More will be said of these specific features later. What should be noted, however, is that none of these features limit the action of the pilot. The 777 design utilizes envelope protection in all of its functionality rather than envelope limiting. Envelope protection deters pilot inputs from exceeding certain predefined limits but does not prohibit it. Envelope limiting prevents the pilot from commanding the airplane beyond set limits. For example, the 777 bank angle protection feature will significantly increase the wheel force a pilot encounters when attempting to roll the airplane past a predefined bank angle. This acts as a prompt to the pilot that the airplane is approaching the bank angle limit. However, if deemed necessary, the pilot may override this protection by exerting a greater force on the wheel than is being exerted by the backdrive actuator. The intent is to inform the pilot that the command being given would put the
airplane outside of its normal operating envelope, but the ability to do so is not precluded. This concept is central to the design philosophy of the 777 Primary Flight Control System."

It seems to have the benefit of computer assisted warning - but not interference.

It's a different approach certainly, and one that seems to work just as well. As I'm sure you're aware, you have to put the A320 a fair amount out of whack attitude-wise before the non-alpha related protections kick in.

The point I was making (albeit not very clearly) was that those early nay-sayers were adamant that an entirely computer-controlled flight control system could not be trusted under any circumstances, and were labouring under the misapprehension that the 777 didn't have one.

Hi,

DW:
Seem's to me that it will be better if:
An entirely computer-assisted flight system controlled by the pilot ....

Which is true of both Airbus and Boeing's FBW systems. Unlike Boeing, Airbus's system does have hard limits, but I'd be prepared to bet you could count the number of times those limits have been encountered in 24 years of service on your fingers and toes. For example, the bank angle limit is 67 degrees - any further and you'll risk a spiral dive and structural damage. Those limits are not there to inhibit pilots, they're there to prevent structural damage and loss of control.

Hi DozyWannabe,
That was the design concept - but sometimes the programmers haven't anticipated every possible eventuality. There have been several heavy landings where the ELAC computers prevented the authority the pilots wanted.
Accident of an Iberian Airbus A320 in Bilbao

I bet the LH A320 crosswind landing pilots wish they had full aileron control once the landing gear touched the runway. I know they should not have been there in the first place (X wind outside limits) - but reducing the maximum amount of available aileron control on touch down doesn't make sense to me.

@rudderrudderrat

If there's a case to change it, they will. It wouldn't have been the programmers making that call in any case, it would have been the aeronautical engineers and the test pilots*. Even the best and most experienced of us codemonkeys have to stick to the specification we're given. A good software engineer will tell you how to implement a specification in the best way, but they can't change the spec themselves.

I suspect that even with a conventional airliner, the engineers would push back on changing a behaviour that manifested itself outside of the operational limits - those limits are there for a reason after all.

[* Just to reiterate the point in general - Software Engineers do not define the limits within a FBW control system. That is the remit of the aero engineers and test pilots. ]

SS
 

Hi,

Why Airbus SAS (former AI) introduced SS in commercial aviation?

23,000 hrs flying in my career has never required a computer to protect my bank limit. Maybe we just need to train pilots, not computer operators. AF447 had computer operators because they rarely flew the aircraft. They just monitored the autopilot. If we continue to use autopilot monitor people to fly we need an absolutely fool proof automated airplane that will never disconnect, no matter what. Is that what we want? Why?

Stall recovery was quite simple, lower the nose and add power not pull up into a deep stall at 38,000 ft. We need qualified pilots back in the cockpit now. Not a 300 hr new guy that will only monitor the autopilot for thousands of hours then become a captain. Does anybody disagree? UAS is an abnormal, not an emergency. Every jet has a checklist to keep you flying with no airspeed. pitch and power for weight and altitude. Or just use the altimiter since it is working.

Which begs the question, "How do you get 'hands on' experience?"http://images.ibsrv.net/ibsrv/res/sr...s/icons/46.gif

This post by Good Business Sense in R&N is possibly food for thought.

Aguadalte,

Wasn't slagging on you, just asking.

Besides, glad you found your logbooks so easily, last I looked for mine I wound up in the dark attic!

The statement is strong I agree. Still, I did not want to write the killer, but just one element in the puzzle that lead to the final result.
I gave my opinion below but we can discuss further.

http://www.pprune.org/6727692-post1024.html

http://www.pprune.org/6812692-post343.html

cause
 

"That’s even worse Lyman, under Normal Law, the system logically thinks it’s time to cancel autotrim by reaching Alpha Prot or slightly above, but when the situation has degraded and Alternate Law is active, the system thinks it’s smart to autotrim all the way whatever the Alpha …" CONFiture.

And to me, the sequence of events should be seen in its logical context... Once climbing, the THS stopped trimming @ ~3 degrees. That would indicate that NORMAL Law obtained...(overspeed react, a/c?)

At 2:10:22 "....ALTERNATE LAW..." PNF

Once slowed, with AoA increasing, the THS re-invigorated, then the a/c STALLED.

By the evidence, and knowing the controls LAW profile, one could conclude that the a/c was reacting to an actual overspeed at 2:10:07. The STALLWARN could have been triggered by an A0A at or above STALLWARN trigger (bug).

Once slowed by ( 'elevators only' climb, ) the a/c switched to ALTERNATE2 and the THS started cranking again....

Other than ruffling feathers, I have not seen that this could not be so.

In the case of controls LAW, the evidence points to overspeed followed by Law degrade, then STALL, with an active THS.

Not knowing the length of time the a/c takes to suss/WARN Overspeed, or ALTERNATE LAW, it is conceivable the direction and fortitude of the airmass, plus controls input (to include THS), could easily cause: climb/upset/STALL......

In any case, the aircraft seems to be telling us that @ Alpha Prot (STALLWARN) the THS is stopped. And that once degraded into Alternate2 in the climb, the THS starts trimming UP once again.

As above....

[BEA]"The flight control law switched from normal to alternate at about 2 h 10 min 05. The alternate law adopted was alternate 2B and it did not change again subsequently."

The PNF took 17 secs to note and verbalize the LAW change.The aircraft was effectively stalled before the THS passed 5°NU.

@Lyman:
 

This is Lyman's logic!.......very well corrected by mm43:ok:

Compare for differences and similarities:

F-GLZU A340-300 22 july 2011:

Avherald Link

BEA reort (French)

Yes! No wonder the BEA want to know what is being said in these cockpits.:uhoh:

Find the differences to the AIRPROX incident to A340 TC-JDN on 2 october 2000:
The turkish pilots responded slightly earlier to the zoom climb. (Turkish 18s, AF 90s).

Isn't it strange that pilots immediately respond to an overspeed warning, but fail to notice a zoom climb that (if it hadn't been protected) would have brought the airplane close to stalling?

P.S. Another notable difference is a change in the control law, which now leaves high AoA protection mode with SS in neutral during 0.5 seconds when the AoA is below AlphaProt.

mm43:
Yes...but the question is: did they have enough pitch command to put the nose down when the THS reached more than 13º nose up?
(please, before replying that the aircraft could have been manually trimmed by them, please consider that apart from direct law flight sim situations, (done my flight sim check today;)) FBW pilots may have the chance to never touch a trim wheel during their whole life...)

A vs B
 

Hi Dozy,

Have been reading the ongoing back and forth on A vs B, sidesticks vs yokes, FBW control, etc. I know you are knowledgable on Airbus control systems and defend Airbus practices when challenged. So in that there is not much going on relative to AF447, I though it might be good to put some things in perspective, A vs B.

Boeing designed the 767 & 757 aircraft in tandem with the idea of shared cockpit design features so that pilots could obtain a common type rating to operate either aircraft. Boeing continued to selected conventional control systems for both aircraft.

Airbus had a difficult time initially selling A300 aircraft as a replacement for either the MD-10/11 or the Lockheed Tri-Star and decided on a new approach to the soon to be A-320 which would compete against the Boeing 737. For this airplane and those that followed, a FBW system was designed and the yokes used in the A-300 were replaced with sidesticks.

Boeing finally decided to introduce a FBW system for the new 777 aircraft being considered. In the design of the Boeing 777, Boeing changed the way it went about designing aircraft. On previous aircraft, Boeing pretty much selected the design and presented the aircraft to the customers. For the 777, eight major airlines had a role in the development of the aircraft. The airlines were, All Nippon Airlines, American Airlines, British Airways, Cathay Pacific, Delta Airlines, Japanese Airlines, Qantas and United Airlines. It was the first aircraft completely designed entirely on a computer. It was decided to retain conventional control yokes rather than change to sidestick controllers. Along with traditional yoke and rudder controls, the cockpit featured a simplified layout that retained similarities to other Boeing aircraft. However, the FBW concept used was slightly different than what Airbus designed for the A320. The differences are in philosophies.

Airbus designed its FBW jets with built-in protections or hard limits. Boeing, on the other hand, believed pilots should have the ultimate say, meaning the pilot can override onboard computers and therefore built-in soft limits. So here is the issue. Should pilots or a computer have the ultimate control over a commercial jetliner as the plane approaches its design limits in an emergency? There were and are strong arguments by pilots on both sides of the debates. Some pilots are of the opinion that computer protection of the A320 and subsequent designed aircraft are very good whereas other pilots support the Boeing philosophy that they must have the final say in controlling the aircraft. There are valid arguments both ways.

One argument was the Boeing 757 Cali Columbia crash where the pilot didn't retract the speed brakes after receiving a terrain avoidance warning. In a A320 the retraction would have been automatically accomplished by the computers. On the other hand, the A320 Habsheim crash was the result of the pilot going below a 50 ft threshold in which the computers assumed the pilot was trying to land. The plane did exactly what it was suppose to do according to the computers and landed in the treetops. The first five accidents involving A-320 aircraft were the result, in one way or another, of misunderstandings between the computers and the flight crews. Obviously this has improved through pilot training, familiarity with the computer control systems, and refinements of the computer control systems by Airbus. Over time, there have been 50 incidents on A320 aircraft involving "glass cockpit blackout". The most serious occurred on a United Airlines aircraft where half of the ECAM displays, all radios, transponders, TCAS and attitude indicators were lost. Due to good weather and daylight conditions the pilots were able to land at Newark airport without radio contact.

I should point out that the Boeing 777 FBW aircraft have suffered only two hull-loss accidents and 6 other "occurrences" with 1,009 aircraft currently flying. Neither hull-loss accidents involved the FBW system. One was BA's mishap landing at Heathrow (engine problems) and the other was an onboard fire in the cockpit due to a faulty oxygen tube while the plane was at the gate in Egypt, there have been no fatalities.

Personally, I would fly on any aircraft in commercial service that either A or B have produced. Each are different in some respects but both share very good safety records when you take away the outlying causes such as hijackings and deliberate crashes. I flew on A320s back and forth to the West coast of the US last week, uneventful, pleasant flights and a nice comfortable seating arrangement by Delta Airlines.

Just trying to be fair and balanced when it comes to a statement like this:


The merger of Boeing and MD had nothing to do with the commercial aircraft side of the ledger. IMHO, it had everything to do with the military side of the ledger as MD was near or at bankruptcy. The US DOD arranged a "shotgun wedding" to save the production of several lines of military aircraft from folding. At the time, there was no redeeming value on MD's commercial side of the ledger.

Turbine D - a fine post, thanks.