A33Zab

This is correct,
thrust(N1) will be locked untill PF move T/L out of climb detent.
The cyan "Predicted N1(TLA position)" indicator is visible at the outer edge of the N1 scale.
To make the T/L setting in agreement with N1, PF needs to move T/L, as indicated by TLA position indicator, towards the N1 pointer (locked).
This must be done in 1 continuous movement otherwise N1 will travel towards this TLA pos 'bug' once TLA is out of Climb detent.

Forget the transient and trend indications these are only available in A/ Thrust.

henra

Although I don't disagree with much of what you wrote these items warrant some comments.

- Uncommanded climb:
Unsure what you consider uncommanded. BEA note stating that PF commanded NU on the stick prior to this climb points to the fact that it was at least not completely uncommanded although we don't know if it was deliberate (potentially not).
- Evidence of 60-70kts updraft:
I'm not aware of any indication/confirmation that the plane encountered significant updraft.
- Aispeed falling to 60kts:
Almost surely only the measured speed in the pitot due to stall of the airflow into the pitot. AoA during descent + descent rate points towards a horizontal airspeed >120kts until very shortly prior to impact where it decreased to 107kts.
- Flight Control ineffective below 120kts: Im not aware we have any positive confirmation for this although it is reasonable to assume that their effectiveness will be reduced below that speed.

wallybird7

Hi Henra

All Valid questions.

“Uncommanded 7000 fpm rate of climb”. Why I think that it was uncommanded.

When a plane is at max cruising altitude (35,000) which the crew noted because the temps were higher than forecast, and therefore unable to maintain a higher altitude, why would they intentionally try to do it? Why would they climb to a higher altitude when they knew they could not maintain it?

In weather a normal climb should not exceed 1000 fpm. Beyond that you are out of normal control. Without access to the performance charts I can’t believe it would be possible to execute a climb at a higher rate than that.

There was never any indication that the pilot requested a climb nor received clearance for one.

At the same time, all hell broke loose and everything turned off.

What would cause all of these events?

Would entering a severe thunderstorm have anything to do with it?

The meteorological analysis showed evidence of strong up and down drafts of up to 60-70 knots. Which translates into 6000-7000 fpm rates of climb approximately.

Is it possible that the plane encountered a severe updraft that carried or pushed it up to a rate of 7000 fpm? Is it possible that the updraft or severe turbulence pushed the nose down at the same time the plane was gaining altitude? And if so, is it not possible that the trim reacted in the opposite direction?

And with the pitots icing up is it possible that the sensors were providing false info to the computers? Some indicate overspeed, some indicated > 60 kts. Which is which?

Was it the pilot commanding nose up or the trim system itself? Which?

And at the top of the zoom what was the state of the plane? To me in a stall that they never recovered from.

Check the charts to determine minimum control speed. I know if you’re anywhere near 120 kts you better be on the ground. Not at 35,000 feet.

Do we really “know” any of this? Of course not. Nobody does or ever will.

wallybird7

Hi Sensor

As we all now know from earlier posts, no such thing as "stall speed" its all about AoA.

In all my years in aviation there were no AoA indicators flying mostly all the A/C made in the US. It was/is all about Stall Speed.

Take-off speed, V1, Vr, V2, climb speed, and approach speed -- all based on STALL SPEED.
Fighters, Transports, DC-4, L-188, B-737,27,57,67. A-310, L-1011, DC-10 etc

bubbers44

WB, yes we agree no normal pilot would pull the side stick up at that altitude for loss of airspeed indication but according to the BEA report they did. No pilots I know would do it so why did they? The BEA report said they pulled full back, I would like to hear the CVR for why they thought that was necessary. Some day we will hear it. Probably well after the Paris air show.

JD-EE

jcjeant, I have been on a plane when the pilot ordered the cabin crew to their seats and said there will be some heavy turbulence. So a simple "watch out" is not even severe turbulence as I've ridden through going from the ONT to ORL or MSP. I'd not read too much into that warning to the cabin crew.

Now, there is a theory that they saw the storm ahead, thought they had a clear path, and discovered it was blocked at the far end by the grand-daddy storm. This supposes misuse of the radar due to, you guessed it, lack of training.

JD-EE

Lonewolf 50, "If the aircraft "knows it is stalled" but the pilots don't, what are your suggestions?"

If the aircraft is well enough instrumented to detect a stall (an AoS sensor we know it has) then why in the seven circles of Hell is that data not on the pilots displays?

JD-EE

jcjeant, regarding short data recorders it'd be nice if the plane was also wired with a 256 gigabyte auxiliary solid state disk on the FDR and CVR that recorded simultaneously with the survivable disk. That should be moderately recoverable in some crashes and would also have upset (and normal) data virtually for the full life of the plane.

jcjeant

Hi,

You forget to open the "sarcastic mode" ... this can help some who don't clever follow this thread
I suppose it's lesson learned for AF and they are presently installing the system you advocate ...
How do they not have thought of earlier ? ... it is a mystery that adds to all those already mentioned in this thread

sensor_validation

So the A-310 didn't use this same AoA stall definition philosophy?

Technically a stall is about airfoils and peak lift AoA, and in a complex wing shape as the A330 this will not be a sudden transition.

There is a definition of "stall speed" which is the "speed below which the airplane cannot create enough lift to sustain its weight in steady 1g flight", and it is clear to me that the A340 in the airprox event went below this. The AF447 must have been close when almost levelling off @37,500 hence I have previously questioned the timing of stall warnings.

HazelNuts39

The DFDR data released in the BEA Update point in the direction of an initial climb that was commanded but unintentional and unnoticed. The performance charts you refer to show the rate of climb obtainable at constant Mach and CLB power setting. The 7000 fpm was not sustainable, it was obtained at the expense of airspeed.

The Meteo France analysis in Appendix 1 of BEA's Interim report no.1 states (values in knots added by me):
Tim Vasquez' analysis mentions that "updraft velocities of 20 to 40 kt occur occasionally" in oceanic equatorial cumulonimbus clouds.

There is no evidence at all that these updraft velocities existed in the CBs that AF447 was trying to avoid, nor that they were actually encountered.

The 1 g stall speed is obtained at 1 g at the stall AoA. It serves as a reference for operational minimum speeds such as take-off speeds, V1, Vr, V2, climb speed, and approach speed.

The stall itself occurs at the stall AoA, not at a particular speed, but at particular combinations of airspeed and "gee".

DJ77

Hi HN39, re # 452

I too started my career with mechanical controls and I have really nothing against FBW or automation: I enjoyed it during my last 10 years on the 777. Perhaps it is not necessary to call Darwin .

I think you missed my point about what I called “a new way of flying aircrafts”. It’s a way you don’t learn at flight school (at least not yet). I meant the throttle system, the sidestick controller with the mechanical link replaced by a priority pushbutton, the auto-trim and perhaps other minor systems. I understand these choices and see that they were, justified from an engineering POV. Airbus knew they were removing some feedbacks the pilots were used to. They admittedly tried to mitigate the drawbacks but pilots had to adapt. These changes were not introduced to please the pilots. Pilots don’t buy airliners anyway, airlines do.

About the auto-trim, when AoA protection is active, further nose up trim cannot be applied. You must move the sidestick backward to fly at a greater AoA. With neutral stick, speed will return to alpha prot speed. This provides stability. In ALT law, when AoA protection is not available, auto-trim is always active. Chris Scott explained that the THS would move to full nose up when unable to satisfy positive g demand in a stall. Is he wrong?
I am sure you know flight dynamics enough to understand why preventing the THS to trim up as soon as stall warning is triggered could achieve.

HazelNuts39

The function of the High AoA protection is to keep the airplane out of the stall. So it is entirely logical that this function is activated at an AoA that is sufficiently close to alpha-max so as not to hinder normal operations, yet far enough below it to ensure correct operation in all reasonably expected operating conditions. These considerations govern the selection of alpha-prot.

What might be considered illogical are certain aspects of its implementation. The AoA vanes obviously cannot distinguish between the 'mean AoA' and the short-period fluctuation caused by turbulence. These fluctuations can be mitigated by suitably filtering the AoA signal, but that would make it less sensitive and hence increase the necessary margin between alpha-prot and alpha-max. Obviously you wouldn't want to filter out the 'real' increase of AoA that occurs when the airplane flies into the powerful updraft that some posters on this thread theorize about. The designers have made the opposite choice by applying "phase advance", which amplifies the sensitivity to turbulence. I guess you need to ask the designers to elaborate the pros and cons of their choice.

The other illogical element in my view is that alpha-prot law is not cancelled when AoA decreases below alpha-prot. Again I suspect that that choice has been made after intensive investigation of all the "what if"-s, but you have to ask the designers.

Every design involves compromises between conflicting requirements ...

BOAC

- I am aware of all that, but it does NOT explain why the a/c was pitched by the system towards alpha max from level cruising flight - that is the explanation I am seeking. I can see absolutely no logic in this function.

Can anyone?

HazelNuts39

Maybe you haven't fully digested my sentence: Otherwise, what would you propose for the protection to do once it has been activated?

BOAC

- do I take it that you are saying that the 2001 incident was caused by alpha reaching alpha-prot during an over-speed event, then? - I would suggest that a far better solution would have been to have it NOT activate in the first place?

HazelNuts39

BOAC;

I resign.

Lonewolf_50

If I may be pedantic for a moment, the absence of evidence is not evidence of absence.

The range of potential updraft and downdraft magnitudes you present, (28 kt to 60kt, depending) bounds the issue wally is pointing to well enough.

There is no reason to believe there wasn't some vertical movement in the vicinity. The questions are:
a. Did they hit a patch of it at some point during the event?
b. What was the magnitude?

Vertical air column velocity, were it encountered, would vary from about 2800 fpm to 6000 fpm based on the numbers you offer in your response to wally. That cannot account for the 7000 fpm climb by itself. It seems pretty obvious that the "big climb" was an energy tradeoff, if not in whole than at least to a substantial degree. There may still have been some vertical airmass aiding and abetting this excursion.

Take the smaller magnitude, for example. Assume for a moment that the plane encounters a 2800 fpm up draft as the pilot has begun to operate in alternate law. To get the 7000 fpm vertical rate, the pilot contributes about 4200 fpm to that via trading energy for altitude ... that's a hell of a change from level flight.

If the magnitude of air column movement is at the higher end, 6000 fpm, then the pilot (or the pilot and the plane, if that's how one sees this event) contribute about 1000 fpm to that dramatic initial climb toward 37,000 plus.

That climb input fits wally's 1000 fpm standard, doesn't it? Sure, but one still needs to consider the pitch attitude. As I was trained, when penetrating turbulent air in instrumnt conditions, you set a pitch attitude (wings level) and do your very best to maintain that pitch attitude as you get bounced around, until you get to the other side of the turbulent patch of air. The FDR data points to attitude increasing, not remaining at a particular value.

Given that data, it is unreasonable to assert (wally) that an updraft is the sole, or even primary, cause of that initial rapid climb.

That doesn't mean that vertical air movement necessarily had no input. It's an unknown, but there may be some indirect evidence that BEA can analyze to get a better grasp on that.

A gust load of that magnitude from outside would, I think, leave a trace in the g sensed by the flight control system. Whether or not this is a parameter captured by FDR is another matter. I don't understand the FDR well enough to say. I also don't grasp how the flight control system filters such external loads in the various laws.

OK's point on rate and dampening suggests to me that the system does filter of such external inputs, perhaps as part of the stability features.

sensor_validation

If we are still discussing the A340 airprox the only explanation I have is that with A/P off due to overspeed, so flight control in Normal mode, engines put to idle by pilot action to stop the engine cycling. Turbulence caused a predicted AoA measurement above alpha prot and the High alpha protection triggered to keep the AoA down to alpha prot. Also possible the initial pitch up triggered "alpha floor" and the subsequent pitch up was in response to the sudden throttle increase?

Further references - first is by mm43?
A340 - A330 Control: Flight & Laws
Aerospace/Aviation: Airbus, thrust levers, thrust system

Only direct relevance to AF447 is that if the pitch up can be caused by an AoA spike we have possibility of a QF72 style (pitch down) uncommanded pitch up if there was a period on Normal before Alt2? I do not believe QF72 event ever explained - two faults one the generation of spurious spiky AoA data, the second the incorrect use of the bad data - they were waiting for further info from AF447 before finalizing report - hardware not identical, but maybe control philosophy?

DJ77

There is evidence of absence. You can do the energy calculation yourself or look at http://www.pprune.org/tech-log/45465...ml#post6523817