RetiredF4

Think it through again.

The general misinterpretation of DW and some others,

--if stick released or neutral , then the nose would drop---

is not changed by your excellent reference. from your source:
Stick neutral equals no command to change pitch rate and no demand to change load factor, therefore the FCPC´s use the elevators to maintain the flightpath. That is achievable in the non stalled environment, but not in a stalled condition. There i would substitute ( but that is my own interpretation) the word FLIGHT PATH (which was 45° down) with something like ordered flightpath (defined by nose position in relation to pitch and speed).

In the fully stalled condition the FCPC´s on the order of the SS tried to achieve a flightpath, the AC was not capable to give, therefore elevators and THS full NU. At release of the SS to neutral, the FCPC´s would try to hold the last achieved by maintaining 1g or by keeping zero pitch change. Both would lead to the fact, that no change on the elevators would schow despite the stick movement.

Only a prominent and prolonged SS down would move the elevators out of the full ND position and hopefully lead to the lowering of the pitch.

But anywhere in the phase of the pitch reduction, when commanded flightpath had not yet reached the actual flightpath (45°down) and stick would be brought back to neutral, the FCPC´s would order again a NU elevator to maintain this flightpath disregarding the fact, that it was not yet achievable.

No fun at all.

DozyWannabe

@OK465

But autopilot is now a function of FMS rather than being an entirely separate and distinct entity, I did not say "all" A/B use Honeywell FMS either, I said that's where they are sourced.

I use FMC/FMS as distinct from FCU (which is the collective term for the flight control logic units).

HundredPercentPlease

Of course - I am only 320 experienced not 330; I was just highlighting the simplicity of the training.

People can continue to discuss the precise modes and reactions of the aircraft to grossly incorrect inputs, whilst completely missing the point of the accident and how to prevent it in future.

mm43

RetiredF4;
Good point, though not sure that Dozy will ' buy it '

However, .... the Flight Path Angle was recorded as NCD during the period in question, so I believe my original interpretation is probably correct.

Gretchenfrage

It’s almost like a clash of generations.

I see a lot of pilots quite happy with an Airbus system, others not. Be it so.

It is basically a Master and Servant relation. We invented automation to be our servant, to complement what humans are less talented for, but still leaving him the final authority.

Today it seems that there is an almost Orwellian shift. The Master has become the Servant. Humans are forced to adapt to his masters programming and to cover the deficiencies of automation, but at the same time giving up some of their authority.

Both sides now basically do what they are less talented for, all in the name of safety.

A traitorous one.

See you at the next accident thread

philip2412

why didn`tthe pnf took controll?me think it maybe a human factor issue.
the cpt had just giveb controll to the younger pilot.bit when all happened
the pnf assumed what was wrong,but wasn`t 100% sure.
so maybe he thougt,when the cpt returned taht he (the pnf) had just took the first opportunity to take controll because he was angry at the cpt decision.


soory but english is not native language

DozyWannabe

I'll give you that if you can tell me at what point the aircraft and it's systems did *anything* other than what the men at the controls told it to do.

@mm43 : On the contrary, I'm quite happy to be proved wrong - my only concern is that the supposition is coming from someone who has stated on several occasions that he does not believe the Airbus FBW system will be safe until the sticks are interconnected and back-driven by the computer.

jcjeant

Hi,

DW
I think there is one truth in writes of Gretchenfrage ...
to cover the deficiencies of automation
Indeed .. the pitot tubes event was not required by the pilots (they never told to the pitots tubes to be blocked) .. and it's a deficiency of the automation (pitot tube is part of the automation chain and is one of the main data source waited by the automation system to work at 100 % effectively)

windytoo

I think this thread has stalled.

TTex600

Did the pilots order the autopilot disconnect? Did they order the Autothrust disconnect? Did they order the THS full nose up?

This is the Rumours & News, "final crew conversation" thread. Let's talk about the crew actions in that light. Why the airplane did what it did is over in the Tech Log section.

You will say that the pilot ordered full THS nose up. I'll say that he obviously did not understand what his control inputs were accomplishing, but that he thought his actions correct.

I'm participating in this because I want to ensure it doesn't happen to ME or one of my friends. IOW, I'm trying to learn from the deceased crews mistakes, not trying to attack nor defend the aircraft.

Rest assured that at least one Airbus pilot, that would be me, is flying today with far more knowledge of the airplane thanks to you and this forum. With that I'm off to commit aviation.

DozyWannabe

No, no and yes. I was referring to the period *after* autoflight disconnect, but if you want to see what happens when the autoflight tries to control an aircraft with bad airspeed information, you only have to look at the Birgenair 757 crash.

I'm not disagreeing with you, but the million-dollar question is "why?". As I said on the Tech Log threads, my flying experience began and ended with my AEF flights as a teenager and even on that minimal basis I knew that pulling up anywhere near the stall regime is a bad move, and over and above that if something untoward happens you don't start moving the controls until you've got a good idea what the problem is and how to solve it - especially if you have altitude to spare.

You'll note that the post of mine you quoted was addressed directly at Gretchenfrage, who has done practically nothing but bash the aircraft throughout most of his posting history. It's been good having new blood in the discussion, that's for sure.

Have a safe trip!

Lyman

WHY is a philosophical question, and requires patience, and an open mind.

Now, HOW. There is a reason these two words exist, and I suggest that for purposes of understanding, HOW is the one that should drive.

Per mm43's post, I see the possibility for the PF to be "tagging along" to an incremental climb of his own command. It implies a misunderstanding of how his a/c works, imo, but does that matter to HOW? His roll inputs are thought to be a possible PIO. At his first pull up, BEA said the a/c did not immediately respond. If true, that is important, because if the climb was the result solely of PF's command, how like a PIO is it? If the a/c was bobbing on the ascent, each loss of trajectory ND wants a correction NU.
Can a climb be the result of a PIO? Of course. Without VS reads, he may think the a/c is losing the altitude it had acquired, needing incrementally more NU to maintain "Level". If his screen was not dependable, and he was unaware, why should he hand the a/c off?

Tangentially answering Glueball, yes, this a/c can climb without autopilot 3000 feet without manual input. And it has, it's in the record, here.

"You don't start moving the controls until you have a good idea what the problem is...." per DOZY.

If you mean PF first inputs were a blunder, you are dead wrong. The a/p quit, two seconds later he announced "I have controls". That means that any handling to follow will be done by the pilot. At some point in time, that is just the way it is, and it gets done. The record shows that he made inputs clearly demonstrating he wanted to correct a flight path that had deviated from S/L IN HIS OPINION. So stop trying to insinuate one's own alternate into the facts. That has been criticised, validly, as compromising a later Pitch and Power solution to maintaining control.

As stated previously, An autopilot will disconnect for reasons other than Unreliable airspeeds. This thread quickly glossed over that the pilot may have been maneuvering with the assumption of loss of a/p to turbulence induced inability to keep up.

It would be interesting to read how the 330 annunciates specific failures to its pilot, to establish that he KNEW disconnect was UAS, and maneuvered quickly in spite of that. Because in the record, the UAS loss of speeds/ALTERNATE LAW, is announced by the PNF, eleven seconds later.

Too fine a point? How so? This is the absolute beginning of the disastrous man/machine interface that brought her down. IMHO.

MASTER CAUTION. CAVALRY CHARGE. PITCH DARK. TURBULENCE. DUFF SPEEDS. HARPY F/O (imo). (DOES he have a screen?)

OK, heroes, what do YOU do?

Dozy: "After autopilot disconnect" re: the pilot commanding THS NU.
Who wants autotrim in ALTERNATE LAW 2? Raise your hand.

GarageYears

Going back to a previous topic in this tread:

A recent study in the USA indicates that cars fitted with ABS brakes are 65% MORE LIKELY to be involved in a fatal accident, compared to those without....

The problem - driver training. Particularly for those that simply regard the car as an extension of their sofa, jump in and press go, without understanding even the basics of the various systems fitted to their car. The report found that drivers unaware of the function of ABS were RELEASING the brake pedal at the point the ABS kicked in to apply-release-apply-release... the brakes for them, with the end result of significantly extended braking distances, which is the exact opposite of what was expected.

The point? Any system intended to aid a human operate some piece of equipment MUST require the operator to understand every applicable operational aspect of the equipment in question. Otherwise, and as seen above, the very point of the safety system may be invalidated by the human operator.

Does this apply to the AF447 crash? I would say very much so. Just like a driver with ABS, the fact the car has them doesn't avoid the need to steer appropriately, select the correct speed for the road conditions, or turn on the windscreen wipers, any pilot sitting in the cockpit of an Airbus or Boeing for that matter better understand the key operational systems affecting his/her ability to fly the aircraft, and if that includes some less common degraded modes of operation, then especially so, since the point where these kick-in are likely to be the same point that the aircraft is about to go pear-shaped or worse.

Dani

I think we have to let go the thought that the accident of AF447 has something to do with the "laws" of an Airbus fbw aircraft. Why? Because the very same incident happend later again on AF471. The later was in normal law without icing of the pitots (as far as we know at the moment). It is clearly a human issue, mostly because of inproper training. You could discuss for ages (and you actually do) that it has to do with the automatics, and it is so, but we have to live with it as long as Airbus are the way they are. But it's no rocket science to find out how you have to handle an aircraft in such a situation. Many crews have proven it since today and will do in the future.

Lyman

Garage Years

Bravo. Any automation MUST be functionally intuitive, otherwise the reliance becomes strictly on training to the system. That would be fine, except....


where is training to the system, here (447)?

Here is the philosophical fundament of the argument, then. Teaching Flying as a mechanical skill is ok, but it is self limiting, in a regime that will ALWAYS require abstract and intuitive thought, at least periodically.

The impasse is also periodically deadly.

subsonicsubic

Currently watching a good dramatisation of the incident.

Pitot tubes will apparently fail. How about implementing a system where the RAT is dropped in such scenarios and airspeed calculated from its rpm.

Now, before you shoot me down, teh rpm based on altitude / air density could be confirmed and corroberated using a static port.

Hat / coat etc....

notfred

From the article that started this threadI haven't seen anyone address this comment yet. They knew they had lost air speed indication, but from this comment I think they believed that they had a static system problem. This could lead them to discount the altimeter showing the climb and loss of height in the stall. Whilst I've learnt from the AF447 threads that the vertical speed indication is inertial based, is this something that they would know / remember under duress?

exeng

As I understand it the RAT is designed to 'self govern' its RPM so I don't think that idea would work.

DozyWannabe

For a start, it's only the PF who says that (at 2:11:58) - the PNF and Captain make no reference to it at all, so there was no agreement and therefore no "they".

By that point they indeed had a problem with data coming from the static ports, in that the airflow around the static ports was being fouled by the "mushing" stall descent profile, thus causing erratic VS readings (see the DFDR VS profile for 2:11:58), and at that point the reading was more-or-less off the scale.

I'm pretty sure that covered static ports tend to report varying altitudes around a certain range, as in the Aeroperu case. I've yet to hear of a static port failure that resulted in a steady, unwinding altimeter that appears to have been the case with AF447, and I'd hazard the opinion that whether you trust the static ports or not, with an altimeter unwinding at approximately 10,000ft/minute it's better to err on the side of caution and take it seriously!

Organfreak

Could you please tell us which one/where? I'd like to be certain I've seen all of them.