DozyWannabe

As he said, he wasn't being entirely serious - he also said that if an outfit was known to be deficient in training that he would not fly with them (and presumably by extension would not allow his family to fly with them).

The first page of the first thread seems to be largely taken up with discussion of finding the flight recorders using their pingers - are you talking about a different thread?

Lyman

Hi Doze,

But in this case there was no apparent "confusion" over laws. The drop to Alternate was called by the PNF, but the PF either didn't notice or disregarded the call. At no point does the PF question why the aircraft is respnsing the way it is and at no point does he quesiton the fact that the aircraft is clearly not in a protected mode.

What a wild statement to make!

The drop to ALTERNATE was called by PNF seventeen seconds after loss of AUTOPILOT. You claim there was no "confusion"? It is more likely PF thought he remained[/I] IN NORMAL than that he thought the LAWS had changed. Evidence would be his ham handed handling of the a/c in manual. One could say he was merely rusty in ALT, but one could also say he thought NORMAL LAW still obtained for flight, and the ship still responsible for smooth.

If PNF was seriously asking "What was that?" re: STALLSTALL, and PF did not even mention it, why would you say he saw amber "ALT LAW" on the screen? One can make no suppositions in the absence of even basic conversation re: CRM.


As to why he did not question why the a/c was "handling different", are you serious? He didn't know it was "handling" at ALL and I suggest he was unable to discern the simple matter of LAW differentiation on the "fly".....

DozyWannabe

Then he should have been listening.


Just as logically explained by the fact that he had no training or practice in manual handling.

Then - again - he should have been listening

Indeed - we don't even know if "What was that?" referred to the stall warning.

He didn't need to - he should have made small, gentle corrections. Instead he bashed the stick around like he was attempting a crosswind landing. Even in Normal Law this would have brought the aircraft to the edge of stall - those inputs were not only in excess of what was required, but also in excess of what was sensible. Discussion of control laws is a red herring, because it has absolutely no bearing on why those inputs were so inappropriately excessive in the first place.

First and foremost - even if he was in Normal Law, the attitude he was trying to command was completely inappropriate at cruise level.

RetiredF4

Just for the record concerning confusion about Laws, itīs not that easy like DW sees it. Look for yourself and call it simple if you like.

DozyWannabe

@Franzl :

The point I was making was that you don't need to know the specifics as far as Alternate Law is concerned - fly the thing as you would normally fly it, just don't expect the protections to be there.

Is there anything in what you've posted that contradicts the above summary?

RetiredF4

Yes, if you are a pilot and understand, how flying itself and management of abnormal situations is functioning in a human brain in opposition to an electronic gadget designed by engineers you will find some points to think over.

And the main problem is, that you normally donīt fly that thing, but you are being flown while monitoring. Except when the sh*t hits the fan.

Lyman

Doze.

Even in Normal Law this would have brought the aircraft to the edge of stall - those inputs were not only in excess of what was required, but also in excess of what was sensible. Discussion of control laws is a red herring, because it has absolutely no bearing on why those inputs were so inappropriately excessive in the first place.

Early on, we were treated to closeup video of short finals/touchdown, the pilots handling were stirring mayonnaise, yet the camera outside showed the a/c virtually on rails, very impressive, to see such stickwork. Now that means they were: Johnny Ace, able to fly through anything on a rail, or a forgiving a/c, smoothing stck to find a stable path.

All due respect, the stick was excessive, and the a/c shone brightly as a sweetie.
So I am calling mayonnaise on Bonin, him thinking the ship would smooth his "whatever" stickwork into grease....e thought he was in NORMAL, and made no effort to modify his stick inputs to a LAW of which he was unaware?

LAW has everything to do with everything, in those first twenty seconds. it took the a/c some seconds to load up, and go up, and once committed, no one would be able to suss her attitude. So think DIRECT in roll, and ask: Would this pilot, aware of his AL DIRECT ROLL, act as if he was low, and the a/c made him look good? I say no, and so I conclude his handling bespoke NORMAL, with protections. At 2:10:22, when PNF says "....ALTERNATE LAW", I think he was preoccupied, no, I KNOW he was preoccupied. Why? A pilot who does not know whether he is climbing, at Mach .80, or descending, has not thought of which LAW obtains. So my case is based on: STARTLE (agree?), loss of scan, loss of orientation, and loss of the plot.....This in the first twenty seconds, max. The only save was for the PNF to take over. He didn't. The rest is merely interesting, but the game was over quickly. The PNF and later the Captain, failed, for whatever combination of reasons, to call off the jam, send the RightSeat to jump seat, and fly the plane.

DozyWannabe

How many times can I say this before it sinks in? The Airbus FBW system was designed by a group that included pilots.

And you're not answering my question. You've got five failure modes there (you didn't bold "In the event of temporary loss of all electrical control:"), only one of which is relevant to the case we're discussing and two or three of which have never actually cropped up on the line. From a piloting perspective the only significant failure mode that requires a different approach entirely is mechanical reversion (i.e USE MAN PITCH TRIM ONLY).

Knowing what those modes mean is helpful, but not crucial. The only thing that *is* crucial is that outside Normal Law there are no protections and as such, more care must be taken.

That is the fault of the industry, not the aircraft or the systems. Even then what you're saying is untrue unless you take a very narrow definition of "flying" to mean direct surface connection, because the flight computers will do whatever the pilot requests them to do (just as their electro-mechanical/hydraulic predecessors did) unless what the pilot requests threatens to damage the aircraft.

Most of the Airbus FBW pilots on here were of the opinion that the pilot in that video overcooked it slightly.

Law has nothing to do with it - even with traditional controls, high-altitude manual flight requires a defter touch than that at lower altitudes, because there's less air resistance.

You mean apart from the whacking great ADI indicator in front of both pilots?

It doesn't matter. Even in Normal Law those inputs would have been excessive. The crucial factor is that he didn't have any high-altitude manual handling experience.

Nor should he or she have to - that's the whole point of the law degradation process.

But will you still feel the same way tomorrow?

The PNF seems to have been entirely reliant on explicit instructions from the Captain to delineate his responsibilities. Once the Captain was there, rather that suggesting that he had an idea of what was going on, the PNF waited for the Captain to perform his own troubleshooting, which wasted valuable time.

I suspect that even after the final report is published, the HF and psychology experts are going to be poring over this one for decades - in terms of complexity it is every bit as morbidly fascinating as KLM4805.

jcjeant

Hi,

The final report will be certainly interesting (more than about thirty days to wait if the schedule of the BEA is respected)
It certainly will conclude with recommendations to avoid such accidents in the future
The trial has come (when ?) will be equally interesting .. but will concentrate more on what could possibly be done before the accident to prevent such a tragedy
The BEA will say what to do .. the trial will probably tell what it should have been done before and why

CONF iture

Machinbird,

I agree with your general comment made on post #850
But it is still interesting to note that on both events you refer to, each time a pilot by his initial action on the sidestick probably tried to correct an attitude that was at least 3 degrees below what the attitude in cruise usually is.

Also, in the AF447 case, a level bust could be understandable and considered as normal as the memory item for UAS states to initially adopt 5 degrees of positive attitude.

I think the disconnection with the altitude became flagrant just later on, when the FD bars reappeared. The crew just didn’t process the altitude when the PF instinctively decided to follow that unrealistic vertical command.

So I agree, if a guy, on a regular base, was manually leveling off at 350, he would deeply assimilate what is the necessary scanning and which level of inputs on the controls are acceptable.

Lyman

@CONF iture

"So I agree, if a guy, on a regular base, was manually leveling off at 350, he would deeply assimilate what is the necessary scanning and which level of inputs on the controls are acceptable.

Absolutely. One cannot discourage manual control at altitude, and then, after something happens, blame the airman for less than sharp high altitude skills.

RetiredF4

True and accepted. You didnīt understand my comment, let me explain my answer a little bit. A computer is doing a failure analysis in simple terms by comparing things in a yes or no system. Out of more possibilities the system will end at a clear decision based on all information and earlier programming in a sequential evaluation process. You can explain that one better than i can, therefore i stop there.
We human pilots have the tendency to grasp all available information at once and in a sum and draw the nearest conclusion out of the available information, the knowledge and our personal expierience. We start problem solving with the nearest probable cause, which might be wrong. The more possibilities there are, the greater the danger to start with the wrong conclusion.

Back to the different laws, if those are not displayed in detail by the system, the degradation of subsystems is not clearly recognizable in such a human thinking process. Itīs not only normal and alternate and direct as you say, its complicated as the excerpt of the manual in my previous post clearly points out.

See above. Iīdidnt answer the question first, because your thinking into that matter is that of an engineer and not that of a pilot. But you are talking to those here and tell them itīs simple three cases and just do that and that. In hindsight you are right, but in reality the crew didnīt know in what kind of law they ended up, wether they had what kind of protections, wether it was ALt1 or ALt 2 or Alt 2b or whatever sublaws are possible and they didnīt grasp the consequences of this degradation and couldnīt base their decision making on that knowledge.

This statement is again not true. Knowing the mode you are in is not only helpful, it is mandatory at some point of the game, and not knowing of the consequences, of the amount of degradation might become crucial, because not knowing might agrevate an otherwise simple situation. Outside Normal Law there is a big variety of protections available or not available, dependent on the individual failure and the degradation the System deems necessary.

Why are you always talking about fault? Any discussion with you ends in statements concerning blaming somebody. I didnīt mention anything concerning who is responsible for that fact concerning hand flying time, i only stated the fact, and you should accept it as ist stands. If you tell people to fly the aircraft in alternate law like you do it in normal law, then that statement is misleading. In normal law you only do takeoffs and landing and everything is done and monitored by the computers. In alternate law the roll yaw is direct, to point only one difference. The THS trim might work or might not work, and so on. Itīs different to normal flying and it is differnet to manual flying in normal law (if you meant that one with your term normal flying).

When faced with a sudden failure the knowledge of the nature and detail of the failure and the consequences of failures is paramount. Under that view its best to keep things simple, limited to few cases to aid in a quick and effective decision making process.

At the moment it seems to be a graceful degradation (from the engineering point of view as you say), which keeps the system on the highest possible level, but it is prone to misunderstandings by the human monitor and operator called pilot. A gradual degradation with few and easy understandable steps could be easier understood and could reduce the error possibility.

Go back to the previous threads, how much we all discussed here the different laws and sublaws and their consequences to input and output systems (Autopilot, FBW-System, protections, flightcontrols...), how many wrong statements had been written about them, and we all learned from these errors.

But again, i only wanted to clarify, that the graceful degradation of the Flightlaws is not that easy to understand like you say here. My personal oppinion, all others are entiteled to a different one.

AlphaZuluRomeo

For once I do. I do agree with you, Lyman.
Now, if you would admit that the startle factor doesn't necessarily need (as shown is another AF widebody from Caracas, which made it to Paris) another aircraft malfunction (beside those we already know i.e. pitot freezing & consequences), it would be terrific.

roulishollandais, I was sarcastic at one point, but didn't try to show too much cynism. Perhaps a translation issue? My concern was about the idea about asking pilots to fly manually in RVSM space: With current mean flying levels, I'm afraid it will lead to mid-airs (or airproxes, at last) far too often to be acceptable. I would prefer it was possible to have every crew in the airliner world being able to do this (pretty basic, when you think about it), but when it come to be reliable upon a long period of time, and detecting/alerting as soon as reliability cannot be assured, I'm sorry but we must face it : a computer (never tired, never distracted) do that better than a human being.

Old Carthusian

Franzl
It all comes down to training and knowing you machine does it not? This removes any misunderstanding of the laws. Then you throw CRM into the mix and you have an solution

RetiredF4

Absolutely correct.

But when we or the manufacturer comunicate like DW the asessment, that it flies in the gracefull degraded modes like in normal mode, then we will never get the management to improve the training and knowledge to the required standards.

DozyWannabe

That's not what I said, and it's not what Airbus say either. What *I* said was that the Alternate flight law was designed to handle as as closely to Normal Law as possible despite the failures that caused the drop to Alternate.

Obviously it can't be exactly the same because there's a failure.

Lyman

OC I'll put it as simplistically as you have....

Airbus: Oversold, Undersupported. Grasp>Reach.


Doze: "That's not what I said, and it's not what Airbus say either. What *I* said was that the Alternate flight law was designed to handle as as closely to Normal Law as possible despite the failures that caused the drop to Alternate.

In this paragraph is the heart of the flaw in the architecture.....

To wit: "Hide the problem"......

CONF iture

More training especially simulator for those who fly and attempt to assimilate the complexity of the magic machine from Blagnac ... You have my vote !

RetiredF4

How about your statement below? (Bolding by me)

DozyWannabe

Po-ta-to, po-tah-to. You can look at it as "hiding" the problem, but it's really just trying to make the problem more manageable for the person in control.

No-one has addressed the point I made that the inputs made by the PF were wildly inappropriate for the circumstances, whatever the control law. Getting hung up on control laws and degradation distracts from the far more worrying fact that AF (and probably several other airlines) was routinely sending crews up in which two-thirds of the flight crew had no training in high-altitude manual handling.

It's another of those rare occasions where we agree - although I would like to point out that the "machine from Blagnac" is neither magic (in fact it's quite long-in-the-tooth technology-wise these days), nor is it any more complicated than any other airliner designed in that era.

The whole point of the flight-deck design in cost-saving terms was purely related to streamlining conversion training between types - it was never intended to give airlines an excuse to cut back on any other type of training.

@franzl - I think something might be lost in translation. "Fly ... as you would normally" is not the same as "It flies the same as in Normal Law" - the first statement refers to how the human should approach things ("Du sollst") and the latter describes the aircraft's behaviour ("Es verhält sich") - albeit inaccurately.