HazelNuts39

OK465:

Thank you for your post #1392. I understand most of what you are saying, except when you say "a more relaxed SS input". A valid experiment would compare the two cases at exactly the same entry conditions and exactly the same pilot inputs. On 'theoretical grounds' I remain convinced that the resulting airplane trajectories would be identical up to the point where the elevator reaches the stop.

infrequentflyer789

That flys in the face of everything published about Airbus C* control law. It is supposed to be path stable, make an input and then the path will be held until you change it. If you point it up and don't have enough thrust then it will continue to pitch up to try and do what you asked, even to stall. The BEA make exactly that point in the report.

Sidestick force is not relevant because you could have let it return to neutral and it would do the same.

THS is not relevant either, until you reach elevator limit, because the control feedback loops will simply command elevator deflection to achieve the path. If the THS helps then they will use less elevator, if not then they will use more. None of that changes sidestick force as there is no feedback.

How are you programming in a 3deg THS limit in your sim when there is not one on the a/c - are you using a THS-failure setup of some sort or are you physically holding the trim wheel to override the autotrim or what ?

Stall is part of a procedure (outside test flights) ?

Trimming in stall recovery (if that's what you mean) damn well should be part of the procedure (if it's got to the point that pilots need a SOP for stall recovery) and definitely should not be unknown. Failure to trim in stall (and approach to) recovery has killed or nearly killed a lot of people, on bus and other types.

BEA doesn't ask because they have excluded autotrim as not relevant to this accident. I can see why: elevators alone would have got them stalled, kept them stalled (see OG and HN39 posts here, but BEA will have access to similar data), and autotrim did not hinder recovery because stall was never diagnosed (BEA conclusion) so no recovery was even started (BEA conclusion).

If BEA find something relevant but then fail to ask the tough questions of the mfr / regulator then you might validly criticise [actually I think they have on at least one point, but no ones really picked up on it here yet...].

You might also disagree with their conclusions, just as some think that their conclusion that the fin stayed attached is wrong (or at least not sufficiently investigated).

But what you are doing is the same as criticising BEA for not asking hard questions about the composite fin and its attachments - there might well be hard and interesting questions still to be asked in that area, but they aren't for this report because they didn't cause this accident.

infrequentflyer789

Have to agree. Not only did AB countenance it, but they built the normal law protections to counter exactly that possibility.

Alternate laws are a failure condition, and 447 was several failure conditions down to the point that the protections were gone. At that point AB system designers didn't lose sight of the possible pilot behaviour, just that without valid airdata they can't build a system to help.

OK465

"a more relaxed SS input" was a veiled reference to once the desired trajectory is established it can be maintained "hands off".

With exactly the same entry condition, as a practical matter, I don't think maintaining identical trajectories could be accomplished with 'exactly the same pilot inputs' to the elevator limits.

I understand what you're saying, that in ALT LAW with a somehow fixed THS setting, the FCS should 'theoretically' reposition the elevator as necessary, even hands off, to maintain the flight path up to the point the elevator reaches 'full travel'.

In practice, in ALT LAW, if the THS is somehow stopped at any position, when further movement would be commanded, the FCPC treats this as a jammed stab and will only exercise about 1/2 of the remaining elevator movement automatically, leaving the pilot about 1/2 of the elevator movement for manual control.

And of course to make the comparison in DIRECT LAW (non Nz) requires a whole different series of pilot inputs.

So I see what you're saying. I guess it would be correct if one considered the limit of elevator travel to be 1/2 travel.

FCPC's are wonderful things.

infrequentflyer789

FWIW I think you are right too. The key issue is probably how is CONF getting his 3deg THS limit in the SIM session. If he is using direct law then we're not even comparing apples and oranges, more like mayo and fruitcake. Might as well say a 737 would have felt different...

The comparison should be: pitch-alternate law with normal THS behaviour vs. pitch-alternate law with 3deg THS limit. With same sidestick inputs, is the trajectory the same (up to elevator stops) ?

CONF - is that what you are saying you have done in SIM and if so, how ?

CONF iture

Who's talking about C* ?
DIRECT LAW all the way since data are diagnotized as unreliable.
Why keep trying to patch on dubious data ?
Keep it simple and straightforward - No ambiguity.

As said earlier, protections and and autotrim can well wait for the next flight, what's the urgency ?

No. If you follow the thread and the accident patern it was all about further trimming into the stall - Where is it desirable ? Please explain.

DozyWannabe

That was in fact a side-effect of the intent. The reason the hard protections are primarily there is to allow the pilot to make control inputs at the limit of travel without getting into difficulty.

The fundamental aspect of manual control as it applies to the FBW Airbus flight deck is that the pilot makes inputs, observes behaviour and then corrects if necessary. Slamming the stick against the backstop and holding it there runs counter to this methodology no matter what law you're in.

Now - first rule of Alternate is that there are no hard protections. If the AF447 PF had heard the PNF's call of "Alternate Law" he had no business making inputs that aggressively.

Except that Direct Law throws the crew into a situation where the aircraft handles differently in *every axis* compared to what they are used to, which would likely be more risky than the current design.

In this case, what you're glossing over is that while autotrim in Alternate Law may have extended the duration of the stall if the correct recovery action had been taken, the fact is that autotrim will *help you recover* from stall by the simple action of pushing the sidestick forward and holding it there.

You're also avoiding the fact that holding the stick back like that at cruise is a thoroughly inappropriate response no matter what you're flying.

korrol

Yes, inexplicably I did get the flight number wrong. Sorry about that. As I can't possibly live this down I'm now withdrawing from PPRuNe.

mm43

Even on your AF337 flight, the cabin pressure in an A330 on a flight of greater than 2.5 hours duration is set at 7,350 feet.

mm43

That seems counter intuitive. Do you mean other than a manual input to the Trim Wheel?

CONF iture

BIG deal BIG risk Dozy come back when you got some experience.

I have never made a secret that the attitude control was inappropriate.
It is not before the THS was at 13 deg that 'the stick was hold back like that'.
And thank you for reminding that holding back a flight control is inappropriate. I'm sure the PNF and the CPT would have loved to know about it when the PF was doing it ...

DozyWannabe

That argument might hold water if every (or even a majority of) line pilot(s) on here agreed with you, but they don't.

It should never have been held back at all!

The attitude on the ADI would have been a pretty big hint.

mm43

CONF iture

After all the debate over the unknown position of the other pilots SS, I now suspect that you would be happy to accept the "Iron Cross" display of the SS position on the PFD following an A/P disconnect and law reversion in flight mode.

Turbine D

I don't mean to be critical here but why would you find the THS movement disturbing? It is part of every Airbus FBW aircraft and quite successful if recognition and acknowledgment of it, by the crew flying, acts in a reasonable manner. In the flight control operation section of the FCOM, the THS is described in Normal Law as to what it does. Then in Alternate Law (ALT 1), it is implicit, similar to normal law with limited pitch rate feedback and gains, depending on speed and configuration. ALT 2 is the same as ALT 1. Only in direct law does the auto trim stop when manual trim is required and noted by an amber message. The auto function/THS movement is the result of whatever is commanded by the PF in sidestick movement, back or forward.
I've thought about what took place after AP/AT disconnect on AF447. It quickly became an unorganized experiment in flight ending with the comment "We've tried everything". In reality it didn't have to be this way.
From an earlier PJ2 posting in the AF447 Final Crew Conversation:
I don't think it can be said any better. The primary decision point was, "If the Safe Conduct of the Flight is Impacted" For AF447, it wasn't but the crew thought it was and they went set off in the wrong direction.
Aside from this on a more general note, there was an interesting revised book requested by the FAA and assembled by Airbus, Boeing, Flightsafety and various participating airlines and others. It issued prior to the AF447 accident, AF was not a participating airline but distribution of this to their flight crews could have been helpful. If you have not seen it, you can view it by going here:

http://flightsafety.org/files/AP_UpsetRecovery_Book.pdf

CONF iture

That's the precise period when the attitude got to 10 deg ND ... what a hint Dozy !

One Outsider

That is not correct either. The cabin altitude is limited to a maximum of 7350 feet, it is not set to it.

CONF iture

No please - Too many data already on that PFD.

DozyWannabe

And the period during the zoom climb where it was way in excess of what would normally be accepted? In any case, if the ADI is showing evidence that pitch stability is a problem then stall definitely moves towards the head of the table as far as probable reasons go.

bubbers44

Agree totally. If you can't see a yoke move look at the ADI.

DozyWannabe

And let's be honest, it's likely that most pilots tend to look *through* the yoke by default anyway...