DozyWannabe

Neither can the B777, and it would be a very difficult task to do the same in the B767 and B757. I apologise if I misunderstood your intent, but from my perspective it read that the FBW Airbus models were intended to be flown under automation at the expense of the flight crew.

Actually in the Birgenair case, the B757 FMS continued to try to fly it, pulling the pitch attitude up to the FMS limit - in that case on the very edge of stall.

Which they should have been trained to deal with...

You tell me how the system did not allow the AF447 crew full control. The aircraft stalled because the system was programmed to give the PIC everything he or she asked for in the event of a systems failure.

Because the late professor who instructed me on software engineering and reliability (who, it must be said, was a dyed-in-the-wool sceptic) banged the Airbus FBW case into our heads as the first thing he did and - despite the fact he visited Toulouse and told us everything he saw - was holding their feet to the fire until his dying day.

Not all humans - unless you have evidence to the contrary.

And if you honestly think I'm pro- pilotless airliners or that I'm some kind of advocate for full automation then you clearly have not read my posts. What bothers me is the repetition and propagation of press scuttlebutt as fact. Computers are great at the dull and repetitive tasks that keep an airliner on course, but they're singularly lousy at being able to deal with situations outside of the norm.

[PS: Spinmeister? If anything I'm anti-spin! Here's the deal - the idea that the A320 was the first step towards fully-automated airliners? Spin. The idea that the Airbus FBW protections would routinely override human pilot input? Spin. The claim that a certain A320 on display "thought it was landing"? Spin. ]

infrequentflyer789

Yes it is an issue.

At Schipol the 737 autotrimmed into stall and trim was never touched in recovery (nor was thrust properly applied, but had it been the result might have been worse due to trim - see e.g. http://www.pprune.org/rumours-news/3...ml#post4787036 ).

Ethiopian (boeing again...) at Beirut was mistrimmed, manually, into stall and crash.

At Perpipnan they did drop into direct law... and it didn't help. They were trimmed up into stall and never changed it. Had they stayed in alternate with auto trim would they have managed the recovery ? Would autotrim have helped or hindered them ?

Surely only if the stick is held back, which will be fatal anyway in stall ?

On the other hand if stick is pushed forward, autotrim (if trim was NU into the stall) will help to reduce AOA and reduce chance of secondary stall when compared with not touching the trim, no ?

There is a difference between ignoring a line of investigation and investigating, eliminating and not pursuing further. The auto trim behaviour clearly has been investigated, based on what is in the report.

Resources are finite, and I don't think the investigation necessarily has the remit to go further into things that may be interesting but not causative or relevant to the outcome of the accident. There are areas I'd have liked more depth on - but I think they've eliminated them too.

My impression looking at the HF slant in the report is that BEA have clearly focused on (1) rather than (2) from:

(1) why are crews (not just this one) stalling and pulling back in stall
(2) why is the trim following the pilot command when (1) happens


So, are they correct to prioritise (1) over (2) - which does not mean they ignore (2) - which problem needs fixing to prevent crashes ? Which problem will become a non-problem if the other is fixed ?

Clandestino

I am afraid that Australian ATSB has shown utter disregard for your ideas that fix is inadequate and in its final report of upset to A330 near Learmonth has found actions by Airbus quite acceptable.

Do you really think wording of Annex 13, 5.12 was pulled out of thin air?

If it is just my opinion that judicial actions can have adverse effect on aviation safety efforts, how come ICAO listens to me? Perchance it's not just my opinion, eh?

Yes. Who said the BEA's word is sacrosanct? That other crews did react to stall warning, if not in completely perfect manner, at least in the one that kept them flying makes it hard to argue that

...if we accept it to be general appraisal and not directed at one specific crew.

No, I'm not saying they got it wrong, just that it's largely irrelevant. Of what use is any protection if the crew is able to control the aeroplane properly? Alpha prot is used consciously only when terrain is a factor. AFAIK, accident started to unravel at FL350 over the ocean.

Is this serious proposal? WTH is wrong with what we have, except it doesn't work at really extreme angles.

Small snapshot? We have DFDR readouts and CVR transcript that clearly show crews have no clue what is going on and if anything is marginally consistent with CM2 inputs it is normal law GPWS escape. however, it was performed over the ocean, at FL350, when protections were lost. Have a look at page 96 of French report (English is botched here) to see what would have happened if both pilots went catatonic - it's the magenta line.

Another dazzling display of failing to understand the basic aerodynamics. Departing the envelope through lift limit is called stall. Report is full of reference to stall.

It was far more then sufficient. Test pilots say so. Page 93 refers.

What then did cause the aeroplane to decelerate with maximum thrust, if there was no drag? Things we discuss seriously on professional pilot's forum are sometimes unbelievable.

Who cares about the difference. Instrument rated pilots must be able to keep on flying any aeroplane (even Airbus) whether speeds are unreliable or totally lost. That's basic instrument flying.

Main issue is that they induced stall - something no other crew succeeded.

Stalled without realizing it.

Because there is nothing simple about taking the aeroplane into area which was a) never entered during flight testing b) was not required to be entered by certification requirements c) was never entered by the type during its 11 years in service, let alone getting her out of it. While there are no reliable data about exact aeroplane behaviour during recovery, it is certain very large nose down attitude would have needed to be achieved to get the aeroplane flying again.

Well, this time they listened to you:
No. Maintain attitude and add power used to be low level approach to stall recovery. Maybe it still is.

Per company training but I can't really imagine any sensible training department teaching pilots how to use reduced protections as they are overridable. There is just one protection that is supposed to be used deliberately... I've already said that

17.9° is not cruise pitch. Full back stick is not hamfistedness.

Past Clmax.

What a wonderful partial view! It was always active in all other cases and helped maintain control. It would have helped the recovery if nose down was maintained.

If it were true, the word would be "lagging", not "damped".

I find idea that pilot would rely only on speed indication to check whether the aeoplane is out of control pretty frightening (though it might help explain somewhat Birgenair or AF447).

Same reason Boeing does... computer can't recognize it.

You keep on saying that without offering practical solution.

Tough. Well it's his posterior in the aeroplane so there's good reason to be responsible.

It would have rolled the same way trying to change flightpath with 2.5G demand and maintaining 1G?!? Well if it were let go after the climb was established it would roll in the same direction but not with the same speed.

1. You probably meant direct law, ALT has autotrim
2. It is BS perpetuated by some. It is very easy to trim Airbus in direct law using wheel. Fact that you move stick and not just release the pressure as you trim did cause some concern to me when I read the manual. It is absolutely easy to do, folks who keep spouting it's a problem have never tried it.

It flies very nicely without any computer intervetion, with direct stick-to-control-displacement of direct law.

Per design. Big deal.

So were the others. What is the use of protection, anyway? Just watch CONF iture rising to bait.

She can. Why are we bringing opinions that are blatantly wrong into discussion? We are not flying passengers in bloomingly unstable Su-27!

RR_NDB

My role here is limited. I consider first to assertively say what i am convinced, absolutely sure and this must be said. Because was important in the crash.

I am available to explain by voice the "practical solution" to an interested pprune member. Simple, just call me and i can teach what to do.

Actually most EE knows that. Simple DSP techniques used extensively in the industry.

In simple words: We need to perform a DSP of the air (analog output obviously) speed information and:

1) Inform IMMEDIATELY to the crew the UAS onset.
2) Block IMMDIATELY GI to ELIMINATE GO (GIGO played a pivotal role)

In a previous post i mentioned an Airbus SAS paper that shows the opposite: The crew must diagnose. IMO this is a serious error. In many aspects.

So important specially when you do not have redundancy at all. (Important sensors failing SIMULTANEOUSLY).




Degrade a System and wait for the crew diagnose is a COMPLETE ABSURD. The chances could be dim even for a designer acting as pilot depending on behavior of the System.

The approach used is WRONG. And SHOULD be changed:

1) UAS DSP detector
2) Actions by crew/System or System/crew

RR_NDB

Quote:

Same reason Boeing does... computer can't recognize it.

We are discussing here a total failure of an Airbus SAS product.

FYI, a very cheap processor can do it EASILY.

A DSP of "audio" frequencies or less.

DozyWannabe

No - I meant Alternate Law, but the sentence that followed it referred to Direct.

Fair enough - I was working on the assumption that it would be harer without tactile feedback, but if you think otherwise...


I think there's a bit of a misunderstanding going on here. The point I was making is that all modern airliners rely on computers whether being flown manually or on auto these days. If the electronics quit, it's a lot more difficult to fly them.

RR_NDB - a modern DSP will not pass certification on the basis of MTTF. As a backup it's plausible, but as aprimary replacement for pitot tubes it won't work.

TTex600

Really? And you know this how? Are you, were you, an Airbus test pilot? Just how is a line pilot to accomplish this state of control?

DozyWannabe

@TTex600:

Clandestino is a former FBW Airbus line pilot, now flying ATR turboprops. Some clarification should be made in that what he is describing is Direct Law, where the computers pass the sidestick inputs direct to the flight surfaces.

The FBW Airbus models will not fly if the electrics are lost altogether - but this is also true of any other modern airliner you care to name.

jcjeant

Indeed this is a factual account and the limit of what BEA want to say
Conf asked about a analyze ....
I find (like him certainly) that what you quoted from the report is not a analyze ..
But maybe (or likely certainly) the BEA think that a analyze about is not relevant ...

TTex600

So you agree with me. Without a computer, you can't fly an airbus. Again, try and stay on topic. My post didn't say that you can't fly an airbus without electricity, it said that you can't fly an airbus without computers.

I've just checked my A320 FCOM (27.40.5) and it indicates that an ELAC (that's a flight control computer just in case the peanut gallery isn't aware) provides "direct pitch and roll" along with other functions such as normal pitch and roll. If my books are wrong, please give me a link to a better, more in depth, manual. Then tell my Director of Training about it as well, as he seems to think his manuals are airbus compliant.

Edited to add: I'm still waiting for the instructions on how to put an Airbus in direct law. I'd like to try and fly it that way if I ever get assigned to a mx flight.

DozyWannabe

And you'll have difficulty doing so in just about any other modern airliner, because the only jetliner being made with wholly mechanical manual reversion is the B737.

See above.

You're in Direct Law whenever you land.

If you want to force Direct Law in flight, then the only way to do so is to sabotage the systems. I don't understand why you'd want to.

jcjeant

Add Newton to the equation (the apple) and your explanation of the deceleration will be complete (it's not only for artillery matter)

bubbers44

I have never flown an aircraft, mostly Boeings, that wouldn't fly just fine with all generator power gone and 30 minutes on the standby instruments then indefinately to destination once clear of weather. Losing an aircraft at FL350 because the pitot system froze up would not be an emergency. If you follow the crazy procedure of going into a 5 degree pitch up and climb power then you will soon get an overspeed. Everybody knows that. Maybe the PF forgot he went to alt rules and reacted as he did pitching up into an impossible 14 degree deck angle because he thought Airbus said you can't stall it so just pulled back and held it for over 3 minutes.

My personal experience is once you become a monitor of automation you gradually lose your hands on flying skills. I experienced it once going from the MD80 back to captain on the 737 200 which is a baby step compared to the AB. I never let it happen again once in the 757. I kept up my hand flying skills to be safe.

Clandestino

Important is what happened, not what someone is convinced happened. That there exists a real world that we are dealing with in aviation is beyond doubt so flying is definitively no place for solipsists.

In washing machines and refrigerators also. It's not the digital signal processing technology, it is the algorithm what matters! No one has so far come with reliable algorithm to automatically detect and deal with UAS nor you have made any practical suggestion. Just wishes.

Cheerfully ignoring that any aeroplane equipped with fault warning system in service today, when involved in UAS, will bombard the crews with undue warning messages.

So you have noticed way things work and developed own theories about it after the system was certified and successfully flown for 20 years? What prompted you to notice?
OK - it still holds it's not difficult for someone aware what he is supposed to do.

I am also not immune to buying BS occasionally. Good thing I had a chance to try it myself so rectified my opinion through personal experience.

Exactly - they are not relying on computers to keep them airborne, they need it to deal with precisions of RVSM, RNAV and CAT3 but they can be safely flown without them. So what if every computer go tits-up, aeroplane is still flying and it's pilots job too keep it so until safe landing.

Direct law is extensively practiced on initial type rating training and later it often is a part of sim refreshers and checks.

1. Actually it was ATR - A320 - Q400 but even if I were 12 year old Justin Bieber fan, it would have absolutely no impact on the assertions I've made here. Ad hominem is logical fallacy no matter if we use it to discredit or support the claim.

2. As these are anonymous forums the origins of the contributions may be opposite to what may be apparent. So if someone makes a claim he is a pilot of this or that and then goes on making absolutely false claims about the equipment he is supposes to fly on, it absolutely doesn't imply that the pilots of said equipment are not acquainted with their mount and that we should blame the manufacturers for providing inadequate manuals.

1. direct law is not to be activated deliberately. Failures that will cause degradation to it are comprehensively listed in FCOM
2. mx flight programme covers way to achieve direct law, if checking the handling in it is required.

Actually, it is only in rollout.

Let me try it this way: everybody knows that, except the test pilots developing the procedure, authorities approving it, instructors teaching it and pilots knowing their procedures. I think you should really talk to them about their fallacy. Bring some arguments along.

Great. Now we're not just creating our own personal aerodynamics but even our personal physics. What do you believe was the reason aeroplane did not accelerate when it started falling? While gravity did help curtailing the zoom climb, original argument was directed at the notion there's not much drag when aeroplane is stalled.

So that's it. He was aware what happened andof the procedure to be applied but unaware that he was flying above acceleration altitude despite taking off four hours earlier. Plausible? I don't think so.

rudderrudderrat

Hi Clandestino,
If you are going to be picky, during roll out, you are actually in a sub mode of Direct Law called Ground Mode.


It's NOT the same as the in flight Direct Law: FCOM 1.27.30 P6.
"With the aircraft in the clean configuration, the maximum roll rate is about 30° per second.
With slats extended, it is about 25° per second."

Ground mode aileron deflection, when above 80 knots, the effectiveness of roll control, ailerons and roll spoilers, is reduced by the half (e.g. aileron deflection limited to 50% of maximum deflection).
See, Report: Lufthansa A320 at Hamburg on Mar 1st 2008, wing touches runway in cross wind landing

Easy for the unwary to get caught out.
Simples?

RR_NDB

Hi,

Clandestino

In a K.I.S.S. approach:

1) An extra resource capable to inform UAS onset would be useful?

2) Block garbage to enter System would be better?


The crew processed adequately the surprises they received?

I would prefer to be informed immediately by an UAS detector resource than to be bombarded by the results (of the UAS) in the System.

False positives would be a problem? IMO, not.

My rationale is:

I donīt like the approach to diagnose UAS by System output. As showed in an Airbus SAS paper mentioned in an earlier post. This can be improved and in AF447 case seems very important.

Thanks to mention Solipsism. Rich issue. Maybe could be applied to Laser based AS sensors. Definitively not here in processing analog information coming from current technology. This is real long time ago.

I started my career using vacuum tubes (big ones in NDB, HF gear, etc.). In the Analog world.

The inputs i prefer (for my decision making) are analog signals. Not garbage processed by Systems.

Sadly, the (non adequately trained) crew of AF447 were:

Surprised by the System output and worse: Had to process garbage generated by the System.

And failed in their decision making. Actually, worse than that, never realizing (in time) what was going on...

bubbers44

I don't think the test pilots that approved the 5 degrees nose up and climb power tried it with blocked pitot tubes. The trapped pitot pressure and the decreasing static pressure would cause an indicated airspeed increase and at some point an overspeed warning.

HazelNuts39

bubbers,

In the case of AF447 the pitot drains remained open, hence the pressure was not trapped. I guess the procedure designers looked at the scenario you postulate, and decided that 5°/CLB would allow sufficient time to get more accurate settings from the QRH, or to analyze the situation and ignore the warning if necessary.

Lyman

HN.... Why would the boffins select a climb, at all, in the first place? If one is in that narrow band of a/s, seems like given the record, higher airspeed might be worth the risk, rather than a climb into the lid, a risk of STALL, or power and lift ceilings being passed, as 447 did? She maxed at 38K, and got their in a flash, so why not 0 degrees and no power increase, or even a reduction in Thrust.

RR_NDB

Even in a test environment misleading is dangerous.

For the sake of Graceful degradation, increasing crew chances to act precisely.

Thiels 727 is a good example of what can occur.

Ignore warnings can be lethal. Better to have decent man-machine interface behavior even during transients and anomalies.

This is possible? I think so. R&D must be done to improve current situation.