You got that right. If they hit TOGA or what ever Airbus calls it, the game is over because you can't overide the pitchup with the underslung engines until they get pitch control.

The aim was not to comment on 'avoiding the stall' nor on how to avoid 'roll reversal' (ie do not use aileron!) but merely an observation on what appeared to be a 'doubting' observation about a known aerodynamic fact that using aileron to pick up a fully stalled wing can put you on your back extremely quickly - aileron on a fully stalled wing ALWAYS produces 'roll reversal' - I remain unconvinced that this lesson is correctly taught or understood - and I agree with Mountain Bear that introducing yet another magic computer programme loop to overcome this is fraught with even more danger.

The 'solution' to our problems lies in flying ability and less reliance on computers to get us out of the mess we have got ourselves into.

EDIT: Before I am jumped on by the theorists, could all please assume the word 'conventional' where 'wings' are mentioned?

stall
 

Hi BOAC.

at the point of stall, or dependent on other factors at a point consistent with approach speed/1.3... the controls should not exhibit any abnormal responses. AoA can be achieved through 360 degrees... ( my rotor blades achieve various AoA's within 0.π to 2.π.... per rev dependent on radial velocity Ω, radius, r/R and translational speed of the rotor system in the airmass, ie inboard section of the span achieves 11/12.π at ψ=180) but the normal regime for a transport is around 0.π to π/12, a very limited sector of operations. From π/2 to 3/2.π the flow is full reversal. As far as adverse aileron roll is concerned there are boundaries of operation where the coefficient of lift v AoA, or "a" slope doesn't decrease with increasing AoA above stall. Equally, drag does not act in a linear manner at extreme AoA's either.

At an α>15 ( π/12) approx to α<30 ( π/6), ie a normal high α > Vs 1g stall α, then it is quite normal to get some control reversal due to the change of the a slope. Discounting the roll spoiler action, the change of section α for the L and R aileron acting differentially needs to be exceeded by the reduced a slope on the section with the downward deflected aileron. This can be avoided by good design of the section and the differential deflections.

With respect to adverse aileron yaw, it is rather moot with a mixed aileron/spoiler roll control, where the spoiler drag exceeds the drag of the aileron pretty much at any α from 0.π to 2.π, ie within a 360 degree rotation of a section.

The A320 accident merely shows a period of behavior within the certified envelope area where there may be some reversal occurring, which is of interest, particularly as an A-PC issue will undoubtedly be held over the pilots head.

To simplify the above, in an aerodynamic stall of a B737 or a B747, the ailerons remain effective at and around the point of break, with normal response, and no exceptional piloting skills needed. The B747 particularly is very nice. Don't do this at home The A320 on the other hand was never tested in this region and the behavior is interesting to this observer, who has examined A-PC issues with this type before. The MD11 ailerons work nicely through stall as well, (but a stall at other than low speed/1.0g will often result in severe damage to the elevators, including delamination of the skins and fracture of the elevator spar... again, don't do this at home...) Having said all the above, which probably amounts to little, the certification process of RPT aircraft is remarkably thorough, but on occasions events conspire to show room for improvement. I would think that the out of trim case for the AI aircraft, and some TBC products still could be improved as we pilots do get to occasionally operate in odd parts of the envelope.

(Probably the greatest control reversal event in recent times was a B737-400, of Adam Air KI 574, which well exceeded the design envelope of the aircraft, and achieved recorded data of roll in the opposite direction of the ailerons/spoilers due to aeroelastic deformation... then it broke up... being operated well beyond the design limits of the aircraft. Shades of Quill & DeHavilland etc...).

Stuff Happens, and the pilot has seconds to deal with it, whereas scientists have years to study it... Unfortunate turn of events led to the event occurring, but the recovery part hardly was a failure of the crew, perhaps I would feel differently if the aerodynamic package for the sims incorporated what we know about dynamics outside of the certified envelope, but at this time the majority of sims are of questionable value outside of the normal envelope, and on occasions negative training (MD11 sim, full aerodynamic stall, ROC 6000FPM for 2 minutes...., B737 stall with full rudder deflection, wings recovered by aileron... B777/B744 50knot Xwind full wrong aileron, wings level to rotate...).

(π= PI, btw.... )

FDR


PS: the problem here from a handling point is the out of trim situation, a bit of "rock and roll" would assist in stopping the attitude from getting higher still due to the out of trim condition.... roll is of interest from a certification point but is far less concern than the out of trim case, or the fact that the failure of AoA probes was not identified and annunciated as a failure to the crew when this condition occurred.

- yup basically what I said.

The airbus was nowhere near 'the point of the stall' nor was it 'within the certified envelope are' nor 'around the point of break'.

yup, but:


“the roll occurring between the stall and the completion of the recovery may not exceed approximately 20 degrees” for level wing stalls. In level wing stalls the bank angle may exceed 20 degrees occasionally, provided that lateral control is effective during recovery.

just playing devils advocate....

this was a wings level entry... it as an approach to stall event.... and it did get to a stall (oops) and the roll exceeded 20 degrees on various occasions in both directions, not completely due to the pilots roll and yaw inputs....

anyway...

No, what the crew did didn't work. What was suggested might have worked, if they had tried it.They used TOGAand they didn't touch the trim either.

The BEA clearly considers that current stall training may be at fault:

Why would pilots with their experience need additional stall recovery training to know that you have to control pitch before TOGA? It is so elementary.

Hi,

But as reminder .. the word of AI (in the past :) )
"Specific training on upsets is not necessary"

jcj - I am not familiar with that quote, but if it is in fact from the 'software junkie' factory it merely amplifies my call for a return to emphasis on basic flying skills and less reliance on automation to save us. If true, a terrifying indictment of the state of mind there.

Stall characteristics
 

Just for context, the 20 degrees limit applies when -
- As soon as the aeroplane is stalled, recover by normal recovery techniques. (25.201(c)(3))
and with -
- The aeroplane trimmed for straight flight at a speed selected by the applicant, but not less than 1.13 VSR and not greater than 1.3 VSR. (25.103(b)(6))

regards,
HN39

Hi,

That's not the first time that FBW and automation are questioned and it will be not the last .....

John T. Halliday: Air France 447: A Cockpit China Syndrome

It seems to me that everyone has become baffled by bulldust and lost sight of the basics. The FACTS are that this crew decided to fly a large jet transport at a relatively low altitude and slow speed. This poor decision was compounded by the pilot flying trimming for a very slow speed. Now, call me a bluff old traditionalist, but could these FACTS just possibly be why the aforementioned large jet transport subsequently crashed.....

Flying at this low airspeed to test their computerized airplane would have probably worked fine at that low altitude but somebody let TOGA take over and put it out of control for the trim setting they had. Boeings I flew never had that problem. You could actually use airmanship to salvage the stall situation. I chose to not ever fly an airbus for this reason. Less money at our airline but better chance of retiring. I retired 7 years ago. Go Boeing!

p51, it wasn’t TOGA (thrust) which was at the root of the handling problem, it was trim.
Thud brings us back to the reality of the accident; it wasn’t the system or the pilot trimming / failing to trim, it was the crew’s understanding of the system / situation which contributed to the accident. This aggravated the poor planning / understanding of what was being attempted.

These issues are not solved by ‘more manual’ flight – basic flying skills. Crews need higher order thinking skills, some of which can be developed and exercised with appropriate manual flight; others require knowledge, application, and preparation – thinking ahead.
Automation might relieve muscular skills and workload, but it is not an excuse to stop thinking.
“Its not pilots hands which get them into trouble, it’s their heads.” (J. Reason)

So now we are going to replace the ever miraculous computer system with the ever present super-human, presumably from the Fourth Reich.

You have a computer system, which can only done what it is programmed to do.
You have a pilot, who can only fly as well as human capacities allow.
You have the interface between the two.

What's the name of that movie..ah yes, "There Will Be Blood." Sooner or later. It's unavoidable.

Hi,

You can think how much you want (and even be Capt Einstein) if the third pilot (the computer) counteract (for any reasons) your thoughtful actions you'll always lose
Indeed .. basic flying skill can't help ....

jcjeant
A big element of truth here. Might well be what happened to AF447.
But on the Perpignan A320, if the crew had been mentally prepared for such wild behavior, I think they could have survived.
Unfortunately, in the context of present Airbus crew training, the phrase "unusual attitude" is supposed to be an oxymoron.

Thnank you PEI for understanding what I was trying to say. Essentially, I was refering to the old adage about a superior pilot using his superior judgement to avoid having to demonstrate his superior skill. P51guy would you deliberately stall (or even fly very slow) a large swept wing jet transport of any type low down? If so, why? It proves nothing (except just how dangerous it is). The truly great Captain Robert Buck said it so much more eloquently than I ever could. While being checked out on the (I think) Constellation by an FAA Check Airman, he was downwind on a blustery day with one engine out when the FAA guy pulled back the power on another engine. Buck knew that things were now getting unnecesarily dangerous, and that being so close to the ground the safety margins were being eroded. He brought the power on the 2nd 'failed' engine back up. When the FAA guy said "you have to do this, it's part of the check" Buck replied "well, looks like I've failed. Let's go home." Buck WAS a Captain. (Incidentally, as he was at the time one of TWA's senior pilots the FAA guy backed down, and they did something else!)

No, I wouldn't have done that. Any normal pilot would have used a combination of thrust with trim control. That keeps the attitude from pitching too much.

Thrust, trim control attitude, airmanship, lack of airmanship, recognizing the failures etc.

All valid points, but it comes to knowing what they were doing and what to expect, if they had ONLY calculated the minimum speed for the conditions of the test like the Airbus Test Pilots do and NEVER go below it, they would have never needed any of that.

They were not prepared and more important, they were not familiar with the test and how to avoid trouble if the test was not successful.

G

Here is where a recognition of training and performance must be addressed. This was a commercial flight, not a test flight, it was flown by demonstrably qualified crew. Perhaps "overqualified". There were conditions at play here that cannot be made comfortable, and foreclose the "old bold" pilot from success. Lack of preparation is rampant, the need to please the customer is an unnecessary companion, etc.

One wonders if it was a low time Captain who was flying, mightn't he decline the manouveur? Something like, "You're kidding, right?". Artificial pressure is dangerous, Always dangerous.

bear

In the meantime, if the aircraft had clearly announced what it positively knew "My AoA probes disagree", the pilots of the day would have never proceeded further with such a test.

What a waste, I hope Airbus learned something too ... ?

Bearfoil:

I strongly disagree with you; this was not a commercial flight. They set out to do tests, to transfer the aircraft from one airline to another. The crew was qualified (or overqualified) for other flights; but the way this flight was conducted, certainly not for this one.

CONF:

They should have never proceeded with the test at that altitude, without knowing what to expect and without calculating minimum speed.

Can't speak for Airbus on changes!

G

That may be true. But airliner annunciations are not, in general, designed for test flights; they are designed for normal operations, and to minimize workload by masking messages which are not expected to be critical - so called "nuisance" or "status" messages. In this case the system thought it was rejecting the one bad value and still had dual redundancy - plenty of margin of safety for a flight where stall approach is an unlikely event. But in this case such a manoeuvre was probability=1. Which throws all the usual risk calculations out the window.

Again, this is why when conducting tests there are different procedures to follow. Perhaps (a/c dependent) you need to check status pages to confirm no faults, or be alert for messaging that would normally be dismissed, or for characteristics that are usually benign but not in these circumstances.

That the ADIRUs didn't deduce from the stuck values that the AOAs were erroneous but valid is unfortunate, but "erroneous but valid" - especially from multiple independent (even if similar) sources - is a horrible failure case to defend against.

As I have said before, there are very good reasons why Airbus specify that this test MUST be carried out at FL140 and not less than 10000 ft AGL. They tried to do it at low level and suffered the consequences.

They simply were not qualified to conduct a test flight and did not understand what they were doing and it bit them on the :mad:.
Any sane test pilot expects any test to fail, they expected to tick a box.

Hi,

Why they will learn now .. when they don't learn before ??

Yes indeed .. what a waste ....
Somebody say's "Retour d'expérience" ? (feedback?)

There really is no point in those posters constantly reminding us that the crew were wrong to try and incorrectly prepared to do this test where they did. Surely that is accepted? May I ask that we stop offering this established fact as the sole 'reason' for the crash?

What this thread needs to focus on is why the 'perfect/infallible' system was not. jcjeant's post now highlights 3 incidences where the root protection system did not function correctly. We should ignore also the 'why's and wherefore's' regarding pressure washing etc. It is apparent that what everyone thought was foolproof was not, and unfortunately the world is not short of fools to test this. How many other anomalies are lurking in the software? When and how will they bite? Will they only bite fools?

We still have 447 'unsolved'.

How to recover?
 

Don't know anything about the Airbus, but please let me ask...


Was this recoverable at any stage prior to the second pitch up?

How?

Was this recoverable at any stage prior to the second pitch up?

It was recoverable right up until they ran out of altitude.
Cancel TOGA power, trim down, normal stall recovery.

The Airbus instructions specify making the test at not less than 10000 ft AGL because recovery could involve losing a lot of height. They tried to do it at low level. Simple.

Lessons to be learned :-
1) Do not allow ground staff to pressure wash sensors.
2) Assess possible improvement to sensor seals.
3) Do not conduct test flights without proper training and briefing.

Are those really the only lessons you can draw? That concerns me.

A "conventional" aircraft without envelope protection would have crashed in the same way if allowed to reach the same configuration. The only difference is that the envelope protection could have saved the day if it had worked.

So maybe one more lesson for ALL pilots, not only Airbus drivers.

Do not allow the aircraft to stall or in any other way exceed the limits of the flight envelope. Bad things can happen. Simple basic piloting skills.
Leave the exploration of limiting corners of the envelope to professional test pilots, they know how to do it safely.

Envelope protection is essentially an add-on to an otherwise normal aircraft which can, in many foreseeable circumstances, mitigate the effects of a bl**dy idiot doing something stupid. There will always be possible unforeseen complex multiple failure modes which the protection cannot handle and the system did not make matters worse, it simply did not make it better. All of which is probably academic because however idiotproof you make any system, the world will deal you a better idiot.

Hi,

A little remind:
Actually (mainly for economics reasons ...) all FBW planes are cruising very near the "coffin corner" and any little "trouble" can put those planes out of their flight envelope.
The AF447 case come to mind (until the contrary is proved)
So ... maybe the FBW planes must be piloted by "professional test pilots" only ??

jcjeant

Most aircraft at altitude are close to coffin corner it does not matter that it is FBW or conventional. You are not supposed to fly FBW aircraft expecting the protections to save you.

As for AF447 we will never know unless they find the "orange-box" and are able to decode it. So don't conclude that it is a FBW design fault, that is just pure speculation on your part without any evidence to support it.

BOAC

Why are you concerned about T A G "lessons" being the only ones?

I thought it was obvious, iceman - there are significant questions in a lot of minds on the philosophy of the 'ultimate protection' system voting out and not telling. They are all here on this thread. If you are to have a 'foolproof' system is has to be.

There can never be an ultimate protection system, the best that can be achieved is a system which protects against the known risks. There will always be rare cases involving multiple failures, in these cases the system should stay out of the way, not add additional risk, and leave the human to sort it out.

Now that we understand the bee in your bonnet perhaps it is time for the moderators to put this long and tedious thread to bed.

To put a finer point on the matter, the FBW/Conventional argument will never die so long as two or more people miss the point. The point is: either, (and both) types perform quite well, and the safety record doesn't suggest anything else. 757s fall out of the sky when a hornet makes a mess in the #1 pitot, and 320s experience a problem with discrepant AoA vanes.

The salient concern, imo, is the interface between the a/c and the Pilots. Whether the downfall is mistaken reliance on pitot or AoA, or a lapse in airmanship, the question remains, How well does the airman know his a/c, and its characteristics of flight in challenging domain, regardless the challenge. Did 447's crew rely on a/p too long?
Is that a human factors/training issue? Here, did the FP perform incorrectly? These aircraft are not sentient beings, they are complex machines, machines that deal out dreadful consequences when operated in less than correct ways. "Blame" as apportioned to the mechanical, seems almost completely unwarranted in any one of the recent outcomes.

Hi,

I then returned again to the "coffin corner"
As you said the greater part of commercial aircrafts are flying very close to the coffin corner.
The reason is mainly for reasons of economy.
But we know very well that when a trouble occurs (FBW or not) the margin is very close to not pass into the coffin corner.
Now they say that everything is constantly put out the flights to take place safely.
But .. airplanes are flying near a danger zone......
It seems to me that there is a paradox .....
But is that the airlines are still airlines?
Are they not become air transport industries (which is different from an airline) and thus are only obsessed with the maximum benefit?

- problem there is you have just torn up the sales brochure.

Trying to condense 48 years of raising my backside from the ground is not easy, but I guess I could say that I have always sought two things:-
1) An aeroplane that lets me know when it is about to let go or has lost the plot
2) The basic flying ability to sort it out when it does.

My flying ability has been adequate, but I have never been easy with an aeroplane that pretends that it knows what is happening but does not - and does not tell me until it dumps in my lap.

Please excuse me if I have become mildly 'allergic' to the indiscriminate use of the term "coffin corner". Please explain what it means, and how near is "very near"? And why "FBW planes" in particular? And since you mention it, how near was AF447 to your "coffin corner"?

regards,
HN39

Hi,

Seem's the "know datas" about AF447 put the flight very near "coffin corner" and the margins for speed or altitude were very thin .. at least.
Wy emphasize on FBW ?
Cause FBW is know for keep the plane flying (more easy than a full manual flying plane) on the limits or to not go of the limits
So as it will be very difficult to fly near "coffin corner" manually .. FBW make it more easy and pilots are more confidents to tease with this dangerous zone.
FBW is not only a help for piloting it's also allowed to build aircrafts differently and use them differently (more economical)
So even the cost of the FBW plane his higher than a non FBW .. air transport industry had understand the economic return of such investissement.
Methink FBW is a good move to more secure flight but this is eroded by the (sometime) use made of FBW by the air transport industry.