AF 447 report out

[roulishollandais]

ACARS

@Emit

difficult to be in the back of marc dubois coming back...
and noting about pitots clogged was written on.the (bad redeable) ECAM.
Robert and cedric heard the printer of the ACARS system and they sometimes took the paper (BEA!!!)

[EMIT]

Prints

Hello Roulis,

I have to scan through the BEA report to find your reference to the ACARS print events.

ACARS and ACMS setup may differ from company to company, but
usually, pilots only get (or make) prints of messages that are addressed to them. Technical messages by ACMS are handled autonomously by the system, no pilot input or checking applies - take for instance the Take-off reports for engine trend monitoring. Only out of curiosity (and boredom) will a pilot ever search for those. The airspeed / Air Data discrepancies that were sent to Air France maintenance were "not for pilot's eyes" . That does not mean that were "kept secret" from the pilots, it is only so that the pilots get their technical indications from ECAM. The indications of what the airplane is doing, is displayed on the Primary Flight Display.
ECAM doen't overwhelm you, you (the pilot) must properly prioritize your tasks, always, all the time.
Could the listing of cryptic messages on ECAM be called a pain in the ass, confusing, etcetera? Yes, but as a pilot, you must be able to cut through the crap, simply keep the nose of the aircraft near the horizon and sort things out properly.
The transcript of the CVR shows, uh, a lot of silence, during the falling piano act of the aircraft, not really a well orchestrated trouble shoot of a nasty, but otherwise simple malfunction.

The ACARS prints that were heard on the CVR may have been weather reports for the enroute alternates, that is the sort of things that pilots use the set for.

Checked the report, found these events.

00:41:50 Noise similar to tearing off an ACARS sheet

01:04:57 Noise of paper being torn


Read the transcript text and you will find it is exactly what I meant: communication with dispatch about alternate airfields, weather, etcetera. During the falling piano act, there is no printer activity listed.

Edited for transcription quotes.

[TTex600]

Originally Posted by TurbineD

Because the perceptions, which are neither right or wrong per-se, lead to understandings are wrong in many instances

Again, if the pilots you guys are writing for do not fly, or do not fly Airbii, why do you care about their perceptions? That was/is the context of my post.

While I'm on my soapbox, and directed at no one and everyone - The self righteousness around here from some who want to exonerate the machine at the expense of the dead is sickening!

If you men really want to correct misunderstandings, you might try to better understand the entire organism of a modern pilot vs focusing solely on the infalibility of his/her machine. Explain how they missed something so obvious as an 11000fpm descent. Explain why the stated on the CVR that they didn't understand what was happening. The ill fated crew was obviously NOT totally incompetent, yet the only reason most pilots on this forum can give for the accident is crew incompetence.

Maybe they just had the wrong perception, maybe they suffered from a misunderstanding, sure that's the answer.

[TTex600]

Originally Posted by Ptimat31

Ptimat31

TTex600,

It is not 'Airbus' Training but 'Air France' training. Please do not confuse the 2. It is based on Airbus recommendations but done in AF premises with AF instructors based on AF SOPs.

"It is based on Airbus recommendations", but it is "not Airbus". Please explain.

Now, back to the Olympics

[A-3TWENTY]

Hi All,

I spent this night in the sim.At the end of the section we still had 30 minutes for free time.
I asked the instructor to level me off at the optimum level for the weight we had , (205T), which was 37000 FT.
It was tested in a new simulator.

I disconected ADRs 2 and 3 and I wanted to force the airplane into stall.

General Impressions:

1.It entered in full stall at around 40thousand feet , well into the red line.

2.To force it into stall I was using a pitch up of between 15 and 20 degrees of pitch and climb thrust.

3.Just before stalling the airplane pitched up by almost byhimself to 30 degrees.

4.It then started sinking R/D between 10 and 15000 ft/min.

5.Then I pushed the side stick full forward and kept it there. NOTHING. Interesting to note that I was maintaining pitch 10º Down , but it never got out of stall.

6.At around 26oooFT I reduced the thrust to get a pitch down moment.

7.It never recovered from stall until the instructor freezed the sim at 4000 feet.

8. Tried the same again. The same happened.

9.Leaving the sim , I met a coleague , very experienced on the A330 who told me thre is a trick to recover from stall. Set manual pitch trim to full down. I didn`t tried it. Maybe next time I have some free time again.

Conclusions:

1. If the sim really reproduces the airplane behaviour , the AF447 was is deep stall and they would never recover.

2.I had never heard about trimming down the airplane.So the crew didn`t for sure.


Your opinions.

[DOVES]

A-3TWENTY
Thank you for your posting, it's been very interesting.
So A330 falls in deep stall.
May I ask you:
-Was there any fuel in Tail Tank? If yes how much?
-Was the stabilizer trim full nose up?
Thank you in advance for your kind response.

[A-3TWENTY]

Hi Dove,

Answering your questions:

I had 50 tones FOB.I didn`t notice how much fuel I had on the trim tank because I asked the instructor to put at 37000 ft. I didn`t climb till there.

So I don`t know if the sim makes the transfer when we are positioned on that level.

The trim:
Probably it was full up , because I kept around 15-20º nose up until it completely stalled.

And completing the latest post.

With the side stick in neutral during the stall it kept a pitch of around 10º in stall.When I pushed it fully forward, the nose came down to around 5-10º NOSE DOWN, but it never got out of stall.

[The Ancient Geek]

All of which is meaningless. The simulator cannot reproduce reality because the conditions are outside of the tested envelope for the aircraft.

[ChrisJ800]

A320
Good post and I assume by disconnecting 2 ADR's that put you in ALT2 Law as per AF447? Its concerning that the system allows full up trim on the THS to be achieved in a short time followed by a stall whilst then not allowing trim to unwind in time to enable a recovery.

I wonder how many manual turns of the trim wheel would be necessary to initiate a recovery and if the automations can be improved to assist in stall recovery?

[A-3TWENTY]

As I told before I wasn`t sure the sim was able to reproduce it , but what you said makes sense.

[RetiredF4]

You only saw, what the sim was giving to you. It did not reproduce the real behaviour of a real aircraft, as there is no reliable and tested data base available.

If you want to know details about it, then read the BEA final report concerning AF447. There is a lot of information in it concerning simulation and simulator behaviour and reliance outside the tested flight envelope.

Concerning the trim the simulator might have been in Alt Law with Autotrim available and might have dropped into Abnormal law during the excessive pitch, where autotrim is no longer available.

[Gulfstreamaviator]

When no data SIM is just a toy

One area of concern that this accident has, at least to me, demonstrated, is that once the aircraft starts to get close to the edge of the known data, and is outside the box, then all guesses as to what the aircraft will do is just a guess....

[Linktrained]

Computers should do what they are told to do.

Lesson 1 in BASIC was to make a computer

PRINT " One + Two = Five "
RETURN

One + Two = Five

There, my computer has a fault ! (Where is the programmer ?)
Even my circular slide rule can tell me how many miles there are in a Kilogramme !

An early computerised N.Atlantic flight plan was much faster than usual. Somewhere 10 Degrees of Longitude had gone missing. ( The 10 Degrees were spotted in good time and replaced.)

[BOAC]

Originally Posted by A-3TWENTY

-1degrees of pitch0

- if you look back through the various threads you will note that many, including me, consider you needed -50 degrees of pitch to unstall. I doubt -10 would do it in time. Are you saying the sim would not GIVE YOU -50 or that you selected -10?

[ap08]

A3TWENTY,

does the simulator produce a log of flight parameters, like a FDR?
if yes, is it similar to the actual FDR data from AF447?

[HazelNuts39]

Originally Posted by A320

3.Just before stalling the airplane pitched up by almost byhimself to 30 degrees.

AF447 showed no tendency to do that.

5.Then I pushed the side stick full forward and kept it there. NOTHING. Interesting to note that I was maintaining pitch 10º Down , but it never got out of stall.

NOTHING? When did it go from 30º Up to 10º Down?

P.S. At what point did you consider the airplane stalled, i.e. what criterion did you use to identify the stall?

[Linktrained]

I was a F/O and also a F/N. At IBM's suggestion I arranged that I and another N/O ( a Straight Navigator) visited American Airlines newish Computerised Flight Planning Center at Idlewild, I think it was in 1964 or '65. They showed us that Coast to Coast they considered 11 different routes and 9 different vertical profiles, Step up and Step down as well as Long Range Cruise. ( It was a total of 99, I remember.) Each was costed. They said that Captains usually took the lowest cost fllght plan. ( I suppose that this was a form of Spread-Sheet presentation.)
They only covered their own routes and the then newish B707.
We were then piston engined. We were long haul charter and might, and did, go anywhere. Often even ground communications were unreliable to many of "our" airfields. ( I went to 80 countries and up to a dozen airfields in some of them.)
Perhaps we were too soon !
A little history.

P.S. Lyman has reminded me that Idlewild had become JFK on 1st. Jan 1964.

[Lyman]

P.S. At what point did you consider the airplane stalled, i.e. what criterion did you use to identify the stall?

From HazelNuts.......


At what point did 447 Stall? What is the relationship of the power ceiling, and the lift ceiling exceedances to the "climbing only due to her momentum"?

I have been trying to establish the cues available to the crew around the STALL.
BEA claim she started her descent without a nose drop, as far as I can tell...

She more or less just started to 'fall', rather than break? This suggests there was no immediate and drastic increase in drag. Certainly there was drag throughout, but the AoA did not appreciably reduce, where is this mysterious and drastic drag arrival? It did not re-orient itself on the airframe, and the g did not shift dramatically, how were they to know they were Stalled?

Wouldn't the STALL more likely be associated with the passing through of the lift ceiling? This happened at what time, can it be known from the report? How long was she climbing on momentum, rather than thrust, and lift?

Linktrained....

Howdy ado... In 1964 if you were at Idlewild, you were actually at Kennedy.

[Clandestino]

Originally Posted by Machinbird

Dozy, The difference clearly is that with the A system the PM has no real idea of the control input history that led to that point of the flight. With the B system, it is available by just resting your fingers on the yoke or visually.

This difference didn't prevent the pilots of conventionally equipped aeroplanes from pulling into stall or Airbus pilots successfully negotiating the loss of airspeed information. However, your claim, which many a PPRuNer would readily agree with, is indeed useful as it reveals not what is wrong with Airbus controls interface but what potentially dangerous habit many a pilot has acquired: checking just the controls movement to see what the aeroplane is doing. Now this might be newsworthy to some: control or thrust levers positions are only demands, which absolutely has to be crosschecked against the instruments to see what effect they have on the aeroplane! That you need to have closed loop feedback via instruments is something trained from first second of instrument flying and is as true on C-172 under the hood as it is in A380 flying through clouds. Ask the last crew of the G-YMMM about the discrepancy between thrust levers position and delivered thrust, you won't be able to ask those who pulled and pulled, yet they could not bring the nose up nor prevent the ground/sea impact.

Originally Posted by TTex600

If the habit to fly pitch and power were as "trained" in Airbus training as is constantly checking FMA for correct mode and values ; and as is a slavish focus on satisfying the ECAM master, maybe we would have no accident to discuss.

1. habit to fly pitch and power is not trained on any Airbus course. It is understood anyone starting the course is very proficient in instrument flying. Preface of any Airbus FCOM refers.

2. when dealing with simple, even multiple failures, ECAM gets it right so following it is well advised, however it is unintelligent computer so it doesn't relieve the pilot knowing very well what he is doing when performing ECAM actions and why he is asked to do that. I suppose "slavish" feeling comes from the pilots too lazy to know their aeroplane properly so can not evaluate whether what ECAM tells them makes sense. I am pretty sure explanation how you figured out there would be no accident to discuss if it weren't for ECAM will be pretty interesting read.

Originally Posted by TTex600

Is it your position that every system is acceptable provided that the operator is "trained"?

No, this is your straw man argument. Not every system is acceptable. Those well designed, tested, certified and field proven stand very good chance of being so. That they can nevertheless turn out to be lethal in unfamiliar hands is just the way our universe is structured. And no, you don't get familiar just by adding the flight hours in the logbook or by hand flying whenever you feel like it. It's about knowing what your aeroplane can and can't do and how to return her into the heart of the envelope, when you have unwittingly chased her towards the edge, or beyond.

Originally Posted by TTex600

Next, you can't have it both ways. Which instruments are primary? Flight directors and FMA's, or attitude and power? The "it could never happen to me" crowd was quick to inform us that the ill fated AF447 crew needed only to fly pitch and power, and now you inform that FMA is important. [/sarcastic bad attitude]

Not much sarcastic but I'll concur it's bad. Let's ignore the last part and assume post was made in good faith. In aviation it often is not case of "either-or but" "and...and...and". Pilot must always recognize when flight director is leading him along the optimal flightpath and so it is to be followed (99.99% of the time), or when it deviates so it must be disregarded. For those doubting the benefits of attitude flying, it is beneficial to remember FD displays attitude command.

Originally Posted by TTex600

It gets a bit annoying to have it implied that I the Bus pilot am an outsider to an airline pilots job, but this is a (basically) anonymous forum so I wont try to convince you otherwise.

"Trust me, I'm a pilot", eh? Ad hominem is still invalid argument, even when it is self-inflicted.

Originally Posted by TTex600

focusing solely on the infalibility of his/her machine.

Nope. This machine was not infallible, none is. It failed in utterly manageable way. Why minor fault resulted in catastrophe was puzzling even for BEA HF experts just piling up theories on top of each other.

Originally Posted by TTex600

Explain why the stated on the CVR that they didn't understand what was happening.

Because they did not understand what was happening.

Originally Posted by A-3TWENTY

If the sim really reproduces the airplane behaviour , the AF447 was is deep stall and they would never recover.

It does not. Covered many a time why it does not. Scrap the rest of the sentence.

Originally Posted by A-3TWENTY

I had never heard about trimming down the airplane.So the crew didn`t for sure.

They could have pushed forward and let the autotrim do the job.

Originally Posted by A-3TWENTY

Your opinions.

Unpublishable.

[RetiredF4]

Clandestino
This difference didn't prevent the pilots of conventionally equipped aeroplanes from pulling into stall or Airbus pilots successfully negotiating the loss of airspeed information.

What a silly straw argument.
Antiskid does not prevent runway excursions completely, therefore those systems are not useful?

Car accidents during braking actions involve more cars with antiskid systems in Europe than cars without such a system. Naturally because most cars on European roads are equipped with such a system.

Clandestino (my bolding)
However, your claim, which many a PPRuNer would readily agree with, is indeed useful as it reveals not what is wrong with Airbus controls interface but what potentially dangerous habit many a pilot has acquired: checking just the controls movement to see what the aeroplane is doing. Now this might be newsworthy to some: control or thrust levers positions are only demands, which absolutely has to be crosschecked against the instruments to see what effect they have on the aeroplane! That you need to have closed loop feedback via instruments is something trained from first second of instrument flying and is as true on C-172 under the hood as it is in A380 flying through clouds.

You name it absolutely correct, and what do you want to tell us?
How is your closed loop feedback ------ demand by pilot via control / thrust levers position versus effect crosschecked via instruments ------ functioning? For the PF it is functioning, as he is the one who is initiating the demand (assumed that the demand is not altered by the automation), the PNF might lack the demand (control position) and has to rely on the effect (instruments) only. The closed loop system is not available to him.

That didn´t go unnoticed during the analysis of BEA and is posted in their final report.

BEA FR 2.1.2.3 Control of the flight path
It would also seem unlikely that the PNF could have determined the PF’s flight path stabilisation targets. It is worth noting that the inputs applied to a sidestick by one pilot cannot be observed easily by the other one and that the conditions of a night flight in IMC make it more difficult to monitor aeroplane attitudes (pitch attitude in particular).