This is what the transducers are supposed to do. Feedback from every control surface is recorded, digitalized and corresponding pressure is applied back to each sidesticks axis. What should have created the legend of "no sidestick feedback" is due to normal C* law feeling of being always correctly trimmed (which is the case).

Since the a/c was almost immediately in "test pilot territory", what role could immediate Autotrim NU have played in a potential recovery? The THS, as takata has said, does not trim UP when g thresholds are exceeded. At the least, it would have provided a more immediate climb, and not required the PF to continue his ever increasing NU demands. Can't think of everything, I guess. That "Zoom" thing. We know that 330 can recover from STALL, it is patent.

Seems if this THS had not held back, the a/c would have STALLED with a great deal more energy in its airframe, perhaps emphatically dropping her nose, etc. The tail drag did them in.

then again, the safety record is exemplary.

It is assuredly coincidence, but when the THS was needed for climb, it was not available. Later, when it was not needed, it came back, and ruined the trajectory. Still later, it prevented a recovery by limiting elevator authority all the way down.

The way I see it

Welcome back, "Bearfoil..."
It was, at first, quite hard for me to parse what you were saying exactly, but now, from a couple of your last post, it becomes perfectly clear again that your style and focus could not be so perfectly immited!
Are we going to chase our tails again on every mentioned subject?

takata

We are a committee. One is a writer, an editor, actually. A retired Pilot, an active PPL, and others, a designer, a mechanic, etc.

If this is illegal, then we say adieu.

If not, when you gather your thoughts, can you respond? You brought it up. The THS is dormant to help in the NU demand. Later, with a climb initiated, it activates and zooms the a/c.

BEA: At PITCH +10, the a/c started to climb, (they mean ascend?)

Are you willing to admit the THS has caused a problem? However unfortunate for the AB Philosophy, (It did as directed), do you have the stuffing to admit that whether the flying pilot knew or did not know, there is a problem when Autotrim interferes in an emergent maneuvering regime?

We have our own moderator, and he cares little for ruffled feathers. You have always seemed fair and passionate. Are you willing?

bon chance

Well, Bearfoil, I have no idea why you (incl. teamwork!) are doing this!
Moreover, me (alone) absolutely never complained about your (multiple) identities!... and behavior to anyone (here, I meant moderators).

I just think that this is a very serious proof of something very wrong with yourself - your many "minds", objectives, agendas.... (including how many other clones?)
I won't complain either as I'm all for open free speech, and it's not my job to moderate you if you (and whoever) are unable to understand that.
My precious...!

FCOM 1.27.10 P3:

"The two controllers are springloaded to neutral, and are not mechanically coupled."

You grant me way too much power. You make more sense when you stick to the topic.

g hobbles? (Preventions) Pilot/Auto "Misunderstanding?" Etc. ?

:ok:

Hi DJ77,
Yes. Which meant that when Captain moves his sidestick, there is no mechanical relation with F/O's position.

Hi PJ2,
How could this ND tendency be introduced if there was no way to change the pressure applied on the sidestick pitch axis?
My understanding is that sidesticks behave differently in auto mode and manual mode, hence will use different circuits. Documentation about those details is lacking in manuals (like many other details: find, for example, the detailed matrix of Airspeed, Alpha, etc; function monitoring).

SS artificial feel.
 

PJ2 & Takata:

There is NO feedback to SS.
Both roll and pitch channel have identical dampers.

Roll channel two rotary spring elements which ensures artificial feel.
Pitch channel two artificial feel spring rods.

The required artificial force in the pitch channel:

http://i474.photobucket.com/albums/r...99/SSForce.jpg

Oh, no, the frigging V/S is going to fall off again.... :{

Hi everybody :-)
 

even though I am not a "flying" person other than sitting at the rear of a plane I've followed this long discussion with enormous interest.
Many questions turn around my head, but I will continue reading here to find some of them answered.
Nevertheless there is one point I would like to have answered: I love the computer, it's technical side as well and have no problem to understand, not all, but most of the problems which might arise. But I know that not only in my family but also among my friends, there are people who dislike using the PC, even though most of them have to use them at their work, in private they tell me after weeks "oh yeah, i have to go and look at my mail" :}

Now the profession of a pilot is first of all to know how to fly a plane, which in itself is already very demanding. But what about a person who loves to fly but has no interest or notion of a computer? He or she just doesn't understand what's it all about as he doesn't care. But in modern flying this is essential. When reading about the reaction of the PF I could not help wondering if he really understood the planes computer and everything connected with it.

I know 320 and 330 work things differently in the flare phase, but I believe the "rolls in a bit of ND bias" is produced by an automatic amount of down elevators, which 'detrim' the aircraft.

Lyman/Bearfoil wrote...


OMG - Bearfoil has evolved into some kind of HIVE-MIND!

Before you shoot off on another wild goose chase concerning the THS, make sure you understand what an "integral" term does in a time-domain control system. (or just look up "integration" on wikipedia). If you understand that, you will understand that the THS position was entirely correct and by-design during the incident.

Hi Jutta - and welcome.

I dont think it's a case of the PF failing to understand the computer (which in this case, outside of autotrim, wasn't actively doing anything to assist his inputs anyway). He seems to have broken a few basic rules of aeronautics which remain the same whether you're in a Cessna, a 747 or an Airbus - chief among those being that you do not pull up into a stall warning*. Other contributing rules include "Do not make large control inputs at cruise level/at the limits of your aircraft's achievable altitude", and the basic rule of CRM, which is "ignore your co-pilot at your peril, doubly so if he or she has more experience than you".

The fact that the Airbus uses a computer to manage many of its functions is neither here nor there anyway - the fact is that if you are certified to fly an airliner and you have people in the back, it is incumbent upon you to know your aircraft's systems well, whether they consist of modern technology or cables, pulleys and bellcranks.

[* - By which I mean the stall warning that was sounding for nearly a minute before the AoA readings became unreliable, during which time he was still pulling up. ]

thanks Dozy for the welcome
 

and your explanation.
What I found very disturbing was the fact that the CDB, when he came back into the CP, did not take his seat immediately. Even if I understand that there are certain rules which have to be followed. But the situation was exceptional, he must have felt that!? Both the captain and other co-pilot put their lives into the hands of a person who was visibly in a lot of stress, mildly expressed. For a laywoman :) this is not comprehensible at all.

With respect, not sure you are correct there - although maybe you have info I do not. Bear came to (more or less) a conclusion in this post:

http://www.pprune.org/rumours-news/4...ml#post6620025

Given the lengths to which he has gone over the past years to find theory after theory to absolve the pilots, that was probably a painful end to the journey. Give the guy some credit for posting that.

I think it unlikely he would be reopening previous arguments after concluding in that way, and even more unlikely that it would be done hiding behind a different name - thin skin was not an attribute of the old bear.

Design
 

sd666

Hmmm.... The "performance as designed" is not in dispute. The THS did in fact, perform as designed. Clear? Because you like many others are stuck on "Design", and 'right' or 'wrong'.

It is, out the gate, quite possible, even likely, that the THS and the PF BOTH performed to "spec".

Fundamentally, the question is this.

The THS lay dormant for a time when the PF was attempting to climb the a/c. It is not programmed to TRIM NOSE UP when the airframe is experiencing g accelerations above, let us say, 1.25g.

The a/c, when the THS came back on line, was climbing, even rotatiing further NU. If unexpected, this could certainly create a "Zoom". A radical PITCH UP, and an unwanted extension of time in aspect (Climb).

So the THS started to TRIM NU from -3 to -13+ without stop, as the PF was still "stick back". Energy, both aerodynamic and ballistic, paid off and the a/c STALLED. Now, the THS is fully NOSE UP, and remained that way till impact. Insufficient ND from PF? Possibly. Stall warning when A/C was close to recover at each ND excursion? So we are told.


Finally, what is the a/c specific performance in Pitch with the elevators free and the THS full NU?

Elevators are sheltered from airflow in this configuration when deflected NOSE DOWN.

They are fully exposed to airflow in this configuration when deflected NOSE UP.

The THS in FULL NOSE UP is in and of itself a very powerful fixed full up Angle of INCIDENCE.

I can frame the question any of a number of other ways. It is meant to be clear and straightforward. Your thoughts?

Please do not feign exasperation to escape answering, as others do.

Again, the THS is not being questioned either as to Performance, or DESIGN.

Right, so one of the takeaways from this incident so far is that CRM (Crew Resource Management - plenty of material online if you need more background info) at Air France was in need of review when this accident occurred.

The difficulties in the Captain re-taking his seat are as follows however:

- The PNF, who seemed to have a pretty good grasp of the situation, yet sadly did not feel the ability to directly intervene was sitting in the Captain's seat. Technically the Captain could have taken the PF's seat, but in doing so would have been putting himself in a position where he wasn't as familiar with the controls as he was in the left-hand seat.

- The two F/Os had been at the controls as the situation was developing and as such, the Captain was unaware of what had led up to the situation - it would have been reasonable for him to assume that the F/Os had better situational awareness than he did at that point (though it was a sadly misplaced assumption)

- The aircraft was pitching and rolling to a significant degree. Exchanging seats at that point would have been difficult and probably would have taken a significant amount of time, and as such could have put them in more danger than they were in already as far as the Captain knew.

Personally I think that the logic for not re-taking a seat at the controls was understandable. What bothers me is the fact that he felt it OK to leave two F/O's, one of whom was fairly junior and probably still a little demob-happy (having just returned from vacation) in charge as the aircraft was transiting a known problem weather area.

[@if789 - I got a very well-written and heartfelt PM from bearfoil just after the report was published, expressing an understanding that human error was probably the largest contributing factor, and a desire to take some time away from the forum - nevertheless he was back on less than 24 hours later having picked up sonething CONF said about control logic and saying that it couldn't have been the pilots' fault again. ]

Hi,

More:
http://www.pprune.org/tech-log/45687...ml#post6630926

Artificial pitch feel

Please enlighten me.

Is it a function of elevator load?

Or is it just a spring rod, no matter if the aircraft is in a stall or at 340 knots?

Thx

Bearfoil,

The fundamental issue is that the PF applied mostly nose-up inputs - hence the THS progressively trimmed up (by integrating his input).

If he had applied the correct stick-forward action in response to a stall, we wouldn't be talking about this at all. The design of the auto-trim is not the cause - the THS acted in response to the PF's input. His inputs were not appropriate. Cause and effect.

In fact, if he had flown pitch and power, it wouldn't have stalled at all.

Indeed,

but what about tactile (pitch) feedback?

Shouldn't a multi-million dollar airliner have a stalled feeling at 100 knots or less?

sdf666

It is the initial handoff that killed this plane, imo. I think that, for whatever reason, the PF wanted to climb, and may have not anticipated a lock out of the Autotrim with a robust Pitch UP. So he kept increasing input (NU), and the a/c started to climb, which lessened g forces, and started the THS TRIMMING NU. This caused a very emphatic PITCH UP, one he addressed with insufficient NOSE DOWN ELEVATOR to reverse the climb angle. In short, because (possibly) he had expected some kind of NL assistance with his pull, and getting none, he kept pulling. When it took hold, he lost his surroundings (SA) possibly, and never caught up with, nor understood, the UPSET handling of his a/c.

I am convinced that some variation of this is what started the problem. What cause dit to be ultimately insoluble is left to others.

Sufficient lack of initial SA, a correct, (though perhaps forgotten) behaviour of the THS, and the rest is coffee table.

Thanks for your response. Finally, look again at your initial statement, separate the sequence of both points, and voila! The initial 'disconnect'

IOW. The THS was NOT TRIMMING up with his first NU inputs. K?

With that collection of expertise, plus a copy of the report, you ought to have figured it out unaided.

C* control law is no mystery, it is a feedback control loop that moves the control surfaces to achieve the demanded movement (by g-load). Autotrim is no mystery either - THS follows elevator to unload it, slowly and with some hysteresis.

At the start of the incident, at cruise speed and alt, the elevator movement required to achieve the demanded climb was small, and the THS moved little or not at all to compensate because it didn't need to. The THS never zoomed the a/c - it started moving only after the zoom climb.

As airspeed bleeds off the control deflection required to achieve the demanded "climb" increases, and as the plane stalls the controls are trying to achieve the aerodynamically impossible - the elevator hits the stops and the THS follows. In the stall. In the stall, because that is what it is being asked to do. The basic control laws have no concept of "stall", just like previous generations of cables and hydraulics - and nor should they, it is far too complex a concept to attempt to build into such low level critical systems.

to JeCe
 

thx for the link. I can agree in some aspects, knowing the other side of the Rhein well ;)

Hey. Absolutely, and in a perfect world. This a/c did not immediately respond to climb input, in spite of cruise speed and controls effectiveness. So, we are left with a need to explain the sluggish response of a highly vaunted "Twitchy" platform. Evidently not? A sluggishness replaced with a near instantaneous and remarkable need for Altitude? After handoff, Throttles were not touched, and the a/c loitered at 34,700 (Read)?

My point is that no one has sufficiently put together a logical continuum of the initial UPSET. There is perpetual reiteration of everything but.

TAG. You're IT.

And of course, you know the diff twixt UPSET and LOC?

Look at the report. The trim does nothing until the a/c is at top of climb and proabbly already stalled. It does nothing because it didn't have to - small elevator movement was enough. Only as it becomes increasingly difficult to maintain the (impossible) climb does the elevator start to move significantly NU and the THS follow. There was no lockout of autotrim. Look at the traces.

Yes indeed that was a strange decision.
Another thing which bothers me is the fact, that the captain asked his co-pilot if he had a licence just bevor he went to rest......... means quite well advanced with their flight.

Is it not normal bevor departure that the team gets at least acquainted and the commander clears important questions like that, bevor boarding?

A while ago I posed the question "In a FBW Airbus how would you sense which way to trim?" to smilin-ed and the answers generated show that many posters don't understand the system.

To answer my own question: "The only sense which can solve the problem is sight - look at the ECAM F/CTL page. If the elevator is neutral with sidestick released the a/c is in trim. If the elevator is displaced then trim in the same direction until elevator is neutral. This will only work with 1g demanded i.e. no sidestick deflection in pitch."

The point I wish to make to smilin-ed is that you cannot feel if the a/c in in trim (except in Direct Law). Autotrim was not the reason for this accident.

We can't tell the tone of voice in which the question was asked, because all we have is the transcription of the CVR. It may turn out that it was just a joke that the Captain was making before he went to get some rest. As is the case in all large airlines, it is possible to get a rostered crew that have never encountered each other before on the flight deck - it's just the way the rostering system assigns personnel.

For example, in the case of one of the best examples of CRM there has ever been (the crash of United Airlines 232 at Sioux City), none of the crew had been personally introduced to Captain Denny Fitch - who was a DC-10 training captain who was "deadheading" on the flight - ever before. But what you hear on that CVR is the trust that all the flight crew, plus their new member, have in each other and the decisions that meant that a lot of people walked away from a crash that was likely to have killed everyone on that aircraft. What I'm trying to say is that in the airline environment, it's a regular occurrence that you'll be working with people that you haven't worked with before - that's not unusual.

Unfortunately, what we're looking at here appears to be at the other end of the scale - a Captain who puts his faith in his two F/Os to manage that leg of the flight, but when things start going wrong it is apparent that at least one of the F/Os (in the right-hand seat) is not handling the situation correctly, and the other F/O (in the left-hand seat) doesn't feel that he has the authority to take over when something goes wrong and waits for the Captain to return, by which time it is effectively too late.

Wuenschen wir doch, als unser Piloten alles klar verstehen koennen. :)

Perhaps the best of the lot, as it was ad lib.

Wir fliegen mit grosstem vertrauen in unsere piloten

Tyro: Tyro, I am well aware the truth in both of these statements. My points are:

1.In my opinion, not having any change in the feeling of the side stick as flight conditions change makes it harder to trim properly. Some people seem to think otherwise, but that is their opinion. This, however is not even a contributing cause of this accident.

2. Autotrim did not cause the accident since the person flying did not even try to recover from the stall which he caused by his persistent nose up inputs. However, if he had tried to recover from the stall it would have been more difficult because autotrim caused the THS to move to the full nose up position. It would have taken more time to attain a nose-down pitching moment than if trim had stayed at the cruise setting.

Smilin_Ed keeps making crushing points from the point of view of a real pilot with his human cargo in tow, well aware of how to ensure their safety to the best of his considerable skill. It makes me happy :) There are still real pilots around, with pride and skill in their craft.

Please don't ask Bear to look at anything, that is akin to work (not even hard work), but inconvenient. The reality of facts gets significantly in the way of today's (or tomorrow's theories...).

The obsession with the THS and autotrim is becoming tedious - the THS did what it was told to do, according to the design of the aircraft. The millions of miles covered by the A330 platform presumably are a reasonable certification of the designs basic soundness - let alone the other Airbus aircraft types of basically identical design.

There is no obsession with THS, it stayed at +13 because no one was flying the airplane, and the Airbus itself was unconcerned with where it landed.

Looking toward the future, how much nose up trim do you really need at FL350?

Are you ever going to drop the flaps up there?

Maybe the control laws need to be looked at in this regard.

Shouldn't some limiting AOA shut down nose up autotrim in Alt law before it gets really ridiculous?

I hate the thought of limiting control authority in some corners of the envelope, but it seems we are moving in a direction where future pilots will need yet more coddling.

Bear is in Alaska, visiting his Polar cousins.

Machinbird, that is a strong point to make. Not new, but strong. Airbus has installed a Trimming system for its elevators (strictly, for the a/c Pitch). There is a reason, a good one. The elevators are a part of a balanced system of demand, and managed stress. As you say, why such an exaggerated system? Possibly, with un"limited" authority, the elevators are subject to large air loads. A movable Stabiliser spreads this stress away from the smaller controls, the elevators.

So why the need for such large excursions in control surface? Why indeed.

Certainly not for authority? At cruise, the deflections are necessarily small, and actually create a danger of damage if they are too large. For added control at Approach to Stall? :ugh:

The RTLU? Ever more important than the Rudder are the elevators. Yet there is no protection at high Mach for these critical surfaces (els)? As has been mentioned many times since AA587, a Rudder can be lost and the flight saved. A vertical Stabiliser if lost is the end of times.

Is the THS indeed a separate approach for control surface protection at high Mach? Then why not fly TRIM? Lock the elevators (a la RTLU), and use the slab to control Pitch. With a dampened and restricted stick at Mach, there will be no Mayonnaise. Likewise ROLL. How can it be so that this dream machine can wobble right and left, inviting a spin, due to PILOT? Especially with such rapid moves?

Always look in the corners, and bring a torch. This investigation has no conclusions as yet. Save those of the Armchair jocks here, evidently.

And those based on nothing more than prejudice and Pride.

Machinbird, the trend is toward more tech, and less pilotage. Shouldn't AB be more attentive to the different "requirements" of the cockpit, and its protections? Is that the question? In the interests of safety and pragmatism, yes of course.

To get defensive (and derisive) not to mention dismissive of decent questions is not only amateurish, it is lazy, even sloppy.

Looking toward the future?? What future? Who needs people? They pay taxes, but you have to feed them. Zero sum game.

I always thought pilots were a cut above, and it still may be so, and what I'm seeing is only the cross-section of the mediocre who bother with forums. I'd hate to think that this is it, in these forums, 3 or 4 real pilots plus a bunch of bull!!!!ters, but it may be so.

Maybe this set of graphs can help a number of points made or questions asked in several posts during this past week or two:

Stall Warning Active and THS Nose UP, from -3 to -13 degrees (max) during the same time interval - see the Gray Area.

http://farm7.static.flickr.com/6090/...5081bb6f_b.jpg

airtren

Perhaps a narrative? In the gray? Because I'm seeing a dogleg at the ELEVATORS/THS Hinge axis. Maybe it's just late.....

How did the aircraft get to be so out of shape with the elevators and THS at virtually the same deflection as where the A/P had them? It Stalled with max Elevator position +/- 2 degrees? And the THS parked at -3 til 2:10:51? Looking for the Dunce cap.......