AF 447 Thread No. 5
 

To make things a little easier to navigate, we'll include a number of links to various spots in the complete thread.

For no reason other than it's a tidy number, we'll try and contain each part thread to around 1000 posts from now given that it is obvious that the thread is going to end up being of epic proportion.

Thread part -

(a) #1 starts here and finishes here. Posts = 3890
(b) #2 starts here and finishes here. Posts = 2537
(c) #3 starts here and finishes here. Posts = 2071
(d) #4 starts here and finishes here. Posts = 1061


Links to the various BEA reports are given below with additional links to the spot in the complete thread where the particular BEA report was issued. If I have missed any of the useful papers, please PM me with the URL and I can include it.

(a) BEA site - French, English
- Report link page - French, English

(b) Interim Report (No, 1) Jul 2, 2009 - English

(b) Interim Report No. 2 Dec 17, 2009 - English
- Update Dec 17, 2009 - French, English

(c) Estimating the wreckage location Jun 30, 2010

(d) Wreckage search analysis Jan 20, 2011

(e) Briefing and associated update May 27, 2011
- Briefing - update French
- Briefing - update English
- Briefing - update German
- Briefing - update Portugese

(f) Third Interim Report July 2011 - French, English


Miscellaneous pertinent links -

(a) Airbus Operations Golden Rules
(b) ALPA FBW Primer
(c) C* and Civil Transports - Cranfield
(d) Longitudinal Flight Control Design - RAeS
(e) Longitudinal Stability: Effect of High Altitude and CG - Boeing
(f) pitot static system performance - USN (Pax River) FTM

Search hint: You can search PPRuNe threads with a filter in Google by using the following search string example -

ths af447 site:http://www.pprune.org/tech-log/

This will search for mentions of THS in the AF447 threads of tech log only.

Just change the THS in the string to whatever you want to look for. This allows one to search for any term or phrase of interest throughout the threads.

Adding the site:URL end part is the magic that restricts Google to only searching in Tech Log.

This filter technique is absolutely wonderful and can be used generically to find things of interest in PPRuNe - appears to work OK in the PPRuNe search function as well.

Excellent work, thanks John - hope you wake up good looking:D

John, first class moderator work, thank you. These threads are now a research tool as well as a "trace of the dialogue". The search tool created by a PPRuNe'r is especially appreciated.

(John, just getting up here on the west coast of N.A. (the 'beauty sleep' didn't work here...better luck!) and I would like to supply the following links to the first thread, (although the real first one (in R&N) got shut down almost immediately and is at http://www.pprune.org/rumours-news/3...0-missing.html. This first thread begins June 1, 2009 and ends on June 4 2009, is 952 posts long and ends at this posting, http://www.pprune.org/rumours-news/3...ml#post4974651.)

The "first" substantive thread, (really the second due the above), is 4598 posts long and was moved to Tech Log some time back and is here: http://www.pprune.org/tech-log/376433-af447.html. It begins June 4th 2009 and ends April 9th 2011.

The first post in that thread is at: http://www.pprune.org/tech-log/37643...ml#post4974708
Last post in that thread is at: http://www.pprune.org/tech-log/37643...ml#post6359812

For the archivists and historians among us! The thread contains some very prescient work.

Reply to Chris Scott, from thread #4:

Chris;
Your previous post on the THS on thread #4, here, explains the THS operation very well and this one connects that explanation with how the THS likely functioned with AF 447. I think its a reasonable explanation of what occurred to the THS and how, after the initial pitch-up.

Missing ACARS explanations
 

Hi All!

This thread has been most interesting to me. Especially on points where the opinions among "experts" differ.
I am still surprised by the fact that BEA gave no hints in the short list explaning where the various ACARS were coming from. One of the last was about faults in Prim 1 and Sec 1, which were earlier assumed to be either showing serious faults of caused by pilot shut-downs of these systems.
Very often it has been assumed in this thread that there was 5 computers involved on AF447, however each of the three ADIRUs also contain at least one each and perhaps also one in each of the Air Data Modules. I am making this conclusion from the australian report about sporadic AoA signals, where it is stated a software update for the ADIRU uncovered an old bug.
I am also thinking about a fixation of the pilots on a single phenomenon as part of the causes, somewhat similar to TMI-II operators failed behaviour.

Regards

Sorry,
but I lost track.

Concerning the "zoom climb", could have A/P and/ or pilots received an "overspeed condition", which in turn traded speed to height, according relevant law?

John T,
Not sure a No.5 link is really the answer....

Unfortunately this is just a forum... without a way to index the various posts into categories like "THS", "Pitots", "Software", "SideStick", "FDR", etc. which would make them easier to "exploit".

Personally just looking forward to the next BEA report... and the way that will put the fox among the chickens, once again.

Missing ACARS explanations?
 

Hi Diversification,

Please, read again BEA interim reports 1 & 2 and specific ACARS chapters. Those reports, including the last note, are all completing each others and what could be explained with all informations on hand at the time each report was printed... was explained.
ACARS is designed for aircraft maintenance, not for aircrash investigations. Hence, some informations in ACARS sequence can not be acertained without access to other sources (CVR, DFDR, or by recovering avionics memories). Nonetheless, at this point, most is already explained or very narrowly conscripted.
What is much more disturbing on the subject is in fact this thread's noise/information ratio around some of the ACARS already explained from day 1 (see Bearfoil's posts still denying pitot and subsequent airspeed issues and consequences on flight systems!).

PRIM 1 and SEC 1 ACARS are only ECAM messages (cockpit effects). There was no "fault" correlated with them that was sent by ACARS.
It could be due either to simple manual reset, or an auto-reset if some fault was detected, by its built-in test equipment (BITE); a single function affected may do that. There is no way to know more without looking into other system memories. If it was due to a manual reset, the CVR won't tell anything if nobody was talking about reseting the PRIM/SEC... The seriousness of this fault is quite improbable as there is 2 other PRIMs and another SEC.

There is effectively five Flight_Control_Computers": PRIM 1, 2, 3 + SEC 1 & 2. They very specific task is to manage pitch, roll, yaw, etc. depending on the various Flight Control Laws, which depend on other imputs (like Air Data, Inertial References, etc.)

Most avionic part of the system may be considered as a "computer". But they are very dedicated "computers" performing some straitforward tasks: ie. An ADM (Air Data Module) task is to digitalise the pneumatic source of a sensor for an ADR unit ; the ADR (Air Data Reference) role is simply to compute the various functions derived from those sensors (Speed, Mach, AoA, barometric altitude) for other "computers" like the FMGC, FADEC, PRIM, SEC, FWC...

I have read the same report and the conclusion was that no "bug" was found on this ADIRU. Nonetheless, something went wrong, twice, with the same unit, but they did not found what it was. AoA spikes filtering was suspected but they could not reproduce it into the lab. Hence, what could have caused it? Was it internal or external? In fact, nobody knows.

At 2 h 10 min 51.......Around fifteen seconds later...........altitude reached its maximum of about 38,000 ft, its pitch attitude and angle of attack being 16 degrees

The recordings stopped at 2 h 14 min 28.

The last recorded values were a pitch attitude of 16.2 degrees nose-up.

So from 10:51 +0:15=11:05 to 14:28, ie 3 minutes +, all the efforts of the crew seem to have had no net effect on NU, but the heading changed. Is this plausible ?

Missing ACARS explanations
 

Diversification;

Just to add another observation to takata's post - at 2h13min the aircraft was descending through FL175, deeply stalled, making uncontrollable roll oscillations left/right.

A similar temporary 'computing' condition may have occurred during the uncommanded pitch-down of QF72 - from the ATSB interim report on QF72:

Based on the released information on the crew actions, it is entirely plausible. The great mystery is why the crew did not take decisive action to unstall the airplane. According to BEA's "3d" depiction of the flight path, the heading changed after point 6, when the airplane descended through FL350 at around 2:11:40. No further information on the heading change is available in the Update. I think the explanation must be sought in the difficulty of controlling the bank angle in a deep stall.

Yes, I understand. But I struggle with it. The clearest interpretaion is that the crew sat for 3 minutes asking for NU and the a/c held its stalled attitude more or less constant through the descent for all those minutes. There are some who may wish to argue that the crew in fact wanted ND but the 'system' frustrated their attempt, but I see no evidence for that intent or that obstruction. Could it be that they just gave up ?

Edit: perhaps as an SLF I should wind my neck in, as they say, but the upshot is the a/c was stalled - clearly and obviously to the a/c systems- but it failed to convey that one vital piece of information to the crew despite all its cleverness and left them guessing for 3 minutes of descent in the turbulent darkness.

without a way to index the various posts into categories like "THS", "Pitots", "Software", "SideStick", "FDR", etc. which would make them easier to "exploit".

I have yet to figure out an easy way to do this. However, the filtered search query will do just this for a given term - not quite as convenient but achieves the aim.

Mr. O

Clearly, BEA believe that what they have witheld holds no danger in current A330 flights, or there would be mitigation; it would be difficult to hide any substantive change to the control system or airframe.

They allowed Airbus to make a statement, and one believes its accuracy, for it will harm the builder greatly if they are found to tell lies. BEA have no duty to "correct" public speculation, for they are comfortable that the information they provided is sound, and no corporate or gvernmental harm lurks.

It is not entirely wise to label what they have done as damage control. Typically, history teaches us that bad news is less disruptive if it is released slowly, and the public has the time both to soften, and go on to other things, (forget).

From the timetable itself, the a/c was in Alternate Law as she climbed. This provides no protection for Roll limiting, but includes AoA protection and direct control for both roll and pitch, aiui.

Right along with doubts about the climb whilst the PF is represented as commanding it, one must suitably indict the a/c for trimming for maximum Pitch UP. What was he THINKING.......

What was she DOING? Certification is mainly a mystery to me; I trust the system to certify a/c to be safe and reliable.

One understands the need for emphatic NU at TO and landing, but in the arena 447 found herself, why doesn't the THS have a LAW LOCK on it similar to the RTLU?

Your question is not so mysterious to me, The PF held back stick because he thought it the correct play. At 10k, his pard thought otherwise, and overrode the PF. "Your Airplane".

The Fourth ACT awaits us later this month?


EDIT. For Mr. O. The a/c systems did NOT know the a/c was stalled, no one did, Why would it protect itself from goofy feet (RTL), and let the Powered slab in back demolish everyone?

Trim is for comfort, when I learned how to use it. Giving any control surface the power to doom the a/c is so............unlike AirBus.

"She did everything asked of her....." Reminds me of what they said post Perpignan.......

Bear, all I can say is that they have my blessing: they were intelligent, trained and wanted to get home.

It is so reassuring to hear you say that, it is in short supply here.

best wishes.

Yes, but my sympathy can't help them. If they did the wrong thing I would be reluctant to hold it to their account with so much software, and so many 'protections', standing guard over them.

"Protections".

The nugget at the core of the Grand discussion. Keep in mind the Protections are to protect the a/c.

Loss of Autopilot in this fbw a/c is not an everyday or mundane thing. It should not be dangerous, and likely was not in this case.

From even the few scraps we were thrown with BEA's note, the first inputs by PF post a/p drop seem innocent enough. They morphed into ever more serious states of flight in very few seconds.

The AB flies "different" out of automatic, not fundamentally, though, and as I see it, the challenge is mounting the horse in full stride with gentle hands. Would this accident have happened if THS "TRIM" was disabled? Doesn't a certified a/c have to demonstrate recovery without trim? What in the World is so much control authority doing in STALL recovery? So much authority that the a/c cannot recover......it says so in the FCOM!!

Simplistically, why does this a/c have anything in addition to NORMAL LAW except bowing out gracefully to let the boys and girls fly it?

Such a needy, dependent, and conflicted aircraft is this one. Has a hard time "Letting Go". Or was it those who programmed her who can't give it up. Control Freaks?

Static (from last page of thread 4)
 

HN39: Thanks also to A337 for the refresher pix of the static ports.

The A330 has an apparent up to 300 ft static port correction at low airspeeds. This is not unusual, but at a greater AOA than stall warning, the static pressure is sure to increase to the point the static ports nearly become pitot ports.

Rather than, "What's it doing now?", the more I read the more I question, Why did they design it that way?"

Well there is the question as to why the crew commanded a climb when 'pitch and power' was what was needed, unless you contend this wasn't wanted. UAS alone shouldn't have caused this loss.

Mr Optimistic.

They commanded a climb because it was proper to do, at the time, airspeeds were fine. They noted "ALTERNATE LAW" eleven seconds AFTER the a/p quit. BEA haven't said where the THS was positioned at loss of a/p. They do say, however, that the a/c was slow to respond to PF's climb input. "At PITCH +10, the a/c began to climb." If the climb had established at AL snatch, what is the protocol for LAW change?

I do NOT believe the PF intended a climb rate at 7kfpm. Nor do I think he wanted necessarily to climb. He pulled back because what he saw and took in told him a roll left and NU was indicated......let's give them that.

At that point, the 340 zoom might or might not be instructive.

Bottom Line? My feeling is that if at a/p drop and then UAS the pilots were expecting, and given Direct Law, there would be no need to be cautious with the stick, to worry about Protections (or depend on them, falsely).

STALLSTALL at his initial pull? Probably a chirp due his unfamiliarity with the a/c response to his hand, and if he relaxed it, (I think he must have) this was the cause of the a/c lagging in climb response.

Greybeard, #18
Interesting point, The higher static pressure would be translated to a decrease in altitude on the display.

However, if the a/c were stalled, the alt would have been unwinding
pretty fast anyway...

- can you support that statement of fact? I would suggest, that having discussed their 'inability' to climb due to the thermal structures of the atmosphere it would have been quite 'improper'?

BOAC +1 ^ :D

How was a "climb" the *proper* thing to do? At all?

Please educate me... given the situation (UAS) surely we ALL agree the proper thing to do is "more or less nothing" (stabilize the aircraft, fly pitch and power, figure out what is going on in a methodical structured manner).

Garage Years and BOAC'

My bad. "climb". "proper". I should have included the quotes, or parents.

The PF at a/p drop ("I have the controls"), Input back pressure.

BEA does not use "climb", nor should I. They state "INPUT".

We do not know the state (assiete) of the airframe at a/p quit, only that it was followed by PF's control ip.

"The a/c at PITCH UP of ten degrees, began to climb."

Did the pilots know at drop they were in UAS? NO. PNF states eleven seconds later, "So we have lost the Speeds, .....ALTERNATE LAW".

We can conclude PNF is reminding PF, or we can conclude he is stating something that has just been noticed.

We can also say that PF may have been trying to capture PITCH and POWER with his movements.

If the a/c had dropped ND a bit, and rolled right a bit in turb, he could have been establishing S/L then intending to P/P.

Remember, the a/c was in weather of some description, if PF had sussed immediately UAS, his first order was to regain S/L. P/P may have been needing an input of Left, NU. He may have been waiting to set N1, until after S/L was attained. Can't do everything at once.

I don't accept that PF ("I have the controls") started hand flying with an utter misunderstanding of the situation. I think his initial assessment was of loss of A/P and then A/T, things happen fast when the wheels are coming off.

My post was not meant to confuse; instead, I think I fell to what is present here, assumptions of events by mutation and groupthink reinforcement.



edit; Anticipating a release in some weeks, I am personally trying to prepare for more data. To start with, there is an embarrassing dearth of evidence to claim anything with confidence.

I still have on open mind, I think stasis in the findings will surprise, and favor a rather standard accident: Too many unusual and unanticipated events happen too quickly, and the a/c and pilot cease a mutually beneficial and safe rapprochement.

No virgins left at the end.

Deep bow, cap off.

Big fan of the F27.

I do have the impression that too many people here are 'parsing' all the 'subtleties' of the English-language translation of the BEA report, trying to 'tease out' information which isn't really there, and isn't there in the original French either.

Being an ancient, and lazy by nature, I haven't bothered yet to put the French and English versions next to each other, except to solve a couple of obvious misunderstandings...

I expect the same thing to happen when the next mid/end July BEA Interim Report is coming out.

As an engineer, I would like to see all the FDR traces - and it's unlikely they will all be put into the report, leave alone on the net....

The CVR might help with a few clues, but I'm not putting the same degree of reliance on it, that some people here seem to do.
"Oh, merde" may be perfectly obvious to the pilots in the cockpit, with one pointing to something visible to all, but no longer comprehensible to those trying to reconstitute the happenings up there, more than two years later.....

Give those people at BEA a chance, and try not to attempt to outguess them too much.

ChristiaanJ

I agree. The essential outcome of this two year ordeal could have been solved (and explained) with the discovery of the Boxes. Witholding data that is in and of itself innocent of bias is traditional, and creates a stage for politics and corporate competition. Except for ass covering and proprietary damage control, I fail to see the need for endless time given to "the Report".

Perpignan, and Habsheim have the stink of controversy yet, as does AA587. The offered report is puffed up and sold as the Truth. The saving grace here, with AF447, is a new vector, extensive and passionate commentary in a very wide audience.

In the last, and for not nefarious purpose, this "Investigative Process" has served to smooth the certain impacts on the commerce that would happen with instant access to the vulnerability of Air Travel to disaster.

Ignorance can be bliss, it can also augment the bottom line.

You mean like attributing made up quotes to the BEA such as
Or
Or continually making wrong statements about the control laws such as
Or just making stuff up such as
Or is it just the general smothering of posts with generalities and ambiguities at a level that normally isn't found outside of the horoscope section of a woman's magazine?

Please continue, but can you back up some of your nay saying?

Start with ALTERNATE LAW?

10 degrees pitch
 

They didn't quite say that, but it may be a literal interpretation of what they wrote in the Update. I have a slightly different interpretation that I would like to throw up for debate. As I read it, one investigator has listened to the CVR and notes the words spoken at 2:10:16 and 2:10:50. Another investigator has studied the DFDR data and gives an account of various observations he has made between 2:10:05 and 2:10:51. He places those observations in a 'logical' order, more or less but not necessarily exactly in the correct time-sequence. For example: Was the nose-down input before or after V/S reached 7,000 ft/min? The reason I'm raising this is not that sentence, but the preceding one: I don't believe these two events occurred in that sequence. If they had, the airplane would be in level flight with pitch at 10 degrees and therefore AoA 10 degrees. At M=0.8 stall warning begins at 4.2 degrees, alpha-max is 5.25 degrees, and the airplane stalls at about 8 degrees. Furthermore, pitch angle does not jump instantaneously from the initial 2.5 degrees to 10 degrees. As soon as pitch angle starts to increase 'progressively', AoA and hence lift increases also, and the airplane starts to climb (give or take small g-variations due to 'chop'). Just can't wait for the real DFDR traces to look for myself ...

Obviously, and especially after reading your last, I am struggling mightily to get a handle on the initial events. You say it so well, and I am ironically satisfied that anyone would single out my struggle for flame.

At some point after origination of PITCH UP, (not the command, the actual), the a/c "began to climb". Does this indicate a gaining of altitude, or arrival at, or transition through, sufficient AoA to enable altitude increase? It enters into question the lack of STALL wrn, but, due Cavalry Charge and MASTER CAUTION, (A/P loss), was a STALLSTALL missed? Logically no, for BEA say the WARNING was "two" in number. implying a cessation of the original WARN at PF's first inputs?

(HN39. Is it at all possible that PF's Right and Left roll commands were in pursuit of an angle steep enough to cause the Nose to fall? It seems too early for such a "desperate step"?)

http://walter.bislins.ch/aviatik/med...n%20Drucks.jpg
the place of the pitots is even sensible, the "static" pressure is not constant over the different places of the fuselage.....the selected place for the pitot is the first place with a static pressure near zero for "normal flight" .....up front is the static pressure higher, behind he is lower, this place is a very good place because the different between the mesured static- and the pitot-pressure shows(indicates) good the airspeed,

with higher AoA the pressure in the static port will get up, (and the correction for the speed) with higher AoA s converted into very difficult.....or impossible

but this is also a chance for an alternativ speed indicating system.....the fuselage itself works like an half-pitot probe...the flow accelerates at the side of the fuselage, and with the different static presure between to different points over the front of the fuselage it must be possible to indicate the speed without a pitot probe!, and the static probe will be not so delicate against ice or ash or insects like a pitot-probe.....

Sorry to say,

PEANUTS.

grity

In a similar discussion earlier, a combination Pitot/Static was suggested in the form of a "Tympanum", a membrane that resembles the skin of a drumhead, flexible, but impervious to water, Ice, or Hornet Nest. The Statics would be arranged about the perimeter, where least deflection of the membrane, and the "Pitot" portion would read at the center, where deflection is greatest. It would take a computer to weight the deltaP by geography, but it is feasible.

I think you answered your question. At that point there wasn't any need for a "desperate step". If they'd wanted the "Nose to fall" they'd pushed the stick forward not pulled back, wouldn't they?

HazelNuts39

I think you are correct, it seems too early. We do not know the situation the PF found himself in. I myself cannot eliminate some form of counter-intuitive aspect of the a/c that baffled the PF's experience and actions.

I cannot accept that the a/c was performing "as designed" whilst the Pilot initiates a radical climb into altitudes that were previously rejected by consensus, a climb that saw 7kfpm, and roll oscillations of plus/minus 12 degrees.

It is possible that all prompts, or most, were not to be trusted (whether they were or not is not relevant),

From a position of balance, given PF's level of skill (high), and the a/c's track record, (impeccable), the two were at odds almost from the outset.

I think the answer will be found in the twenty seconds that bracket the loss of the autopilot.

thanks for your comments, it is an honor to be in conversation with you.

I've been thinking about where BEA is likely to be going in their investigations of the human factors aspect of this accident.

I suspect the words 'cognitive fixation' will factor significantly into their report. (In French of course.:})

In the course of reading up on the subject, I found that there are some simulation tools for predicting this type of problem and quantifying it.
Here is a 6 year old paper on aviation related work in this direction.
http://www.humanfactors.illinois.edu...ulebleiwic.pdf
Now if they were to take this set of models and apply it to the piloting tasks faced by the crew of AF447 that night there might be some clarity to the perplexing question of how they let the aircraft get away from them.

Of course with the CVR, DFDR and possibly ?? the high resolution recorder yielding information, that may be overkill. I think we have teased all the information from the initial BEA note that can be extracted at this point.

Judging by your comments, I take it you've never been involved in an accident investigation, nor have been involved in 'reading' and fully interpreting an FDR recordng. The bias is all yours, I'm afraid....
Perpignan doesn't (company pilots playing "test-pilot" at low altitude without really seriously knowing what they let themselves in for).
Neither does Habsheim which had a pilot with no serious knowledge of the aircraft, getting himself somewhere in the envelope where he shouldn't have been, while switching off some of the systems, then blaming the aircraft and the systems, rather than his own incompetence.
Me262 and Meteor pilots were well aware, that a jet engine at idle needs time to 'spool up'. this jerk obviously didn't....

Curious, what would that achieve exactly ? Based on the info we have (and see the helpful timelines previously posted by others) elevator authority alone was sufficient to get 16deg pitch up and fully stalled. THS simply relieved the elevators.

Had THS not moved, and elevators held 16deg nose up instead, what benefit results ? Only (as far as I can see) perhaps quicker recovery - but THS trims for neutral elevator, so they should, and it appears did, still work if ND is attempted. Unless of course we are into deep stall situation (I would say we don't know if that was the case yet), but in that case THS position quite possibly doesn't matter.

It seems that what you really want is more an attitude-protection law - 10deg pitch up in cruise is silly so sorry dave I can't do that.

But we already have that, in a better way, with AOA & speed protections, but these are lost here - because the planes sensors (or at least some of them) are lying and it doesn't know what is happening with sufficient confidence to apply protections.

Should AOA protections degrade to attitude protections of some sort ? More laws / sub-laws ? More variables combinations to learn and find time to actually be trained on. Does that really help ?