AF447

[lhp]

AP/off

@Willoz469

According to ACAS, the Auto Pilot off is one of the very first messages. SafetyConcern says, the message doesn't indicate whether it's an automatic or manual switch off. Since it was not preceeded by any other material messages, I think it likely that the autopilot was turned off manually. To me, all the messages that follow are just fallout. The real event happened before 02:10 when the AP went off. And since there wereno failure messages before the AP/off, it wasn't equipment failure that made the pilots turn off the AP. It wasn't a VS breaking off, as the hydraulic failure messages would have preceeded the AP/off message.

[Mad (Flt) Scientist]

Originally Posted by lhp

@Willoz469

According to ACAS, the Auto Pilot off is one of the very first messages. SafetyConcern says, the message doesn't indicate whether it's an automatic or manual switch off.

I believe someone stated much earlier in the thread that an ACARS message would not post for a manual AP selection. Which makes sense, otherwise every flight would trigger an AP ACARS at end of flight when you get to min use height for the Ap.

I'm trying to find where it was stated (there's a lot of junk here!) but that the AP was not intentionally turned off seems more logical.

[cesarnc]

Numbers of bodies updated to 24... Good for the families and I assume it's good for the investigation as well...

[Dutch Bru]

Heavy, Scientist

There is indeed a thin line between the clear objective on the Brazilian side to show that they are up to the salvage job so far out on the ocean and the need to protect the feelings of those who are in mourning.

Indeed the Brazilians are (laudably) going out of their way, and not only in the figurative sense. I read tonight in the Brazilian news a report that 8 more bodies were salvaged (indeed upping the total to 24) by the Brazilian Navy, from what is arguable no longer considered "Brazilian waters" (reportedly halfway to Africa east-north-east of TASIL and over 400 km northeast of St Peter and Paul Rocks)

Chapeau!

[Cloud Cuckooland]

lhp (post #782)
The fault reported in the ACARS message (see page 29)
is "AUTO FLIGHT AP OFF" this is generated by an involuntary disconnect - not manual disconnect. In this case it looks like it was caused by multiple ADR disagree and tripping to ALTN LAW.

Dutch Bru:
The portion wreckage showing the Crew Rest compartment labelling - this is called the LDMCR - Lower Deck Mobile Crew Rest and is just as your LH link photo looks i.e in the forward end of the aft cargo (Hold 4).

Im my airline it is used for cabin crew and one of the pilots if 4 pilots are carried. The Captain, when resting, is always in the other rest closer to the flight deck. Interestingly, EK pilots have previously complained in this forum about the distance of the crew rest from the flight deck in case of emergency - (above door 5 in the B777 and mid cabin EY in the A380) - not sure about their A330's. Other operators generally have the Cockpit crew rest close to the cockpit, however you still have the problem of getting through the security door in a hurry when those inside are too busy/distracted to release it. The only airliner I know of that has the optimal crew rest location (bunk room inside the cockpit) is the B744.

[Nakata77]

the tail fin concerns me. it looks like a fatigue fracture across the base of the tail - a perfectly shear split in a perfect line. that is very strange. if the tail fin broke off because of overload then the aircraft would enter a horiztonal spin hence the pitot tubes seeing differing speed indications and eventually the aircraft will throw itself apart. The g forces on the pilots at the front would have been too great to even do anything

[Diamond Bob]

Vertical Stabilizer

I've been suspect of the A300 and by extension the A330 (though I honestly don't know how much different they are), since I saw these photos of the AA587 VS. I just can't imagine a metal attachment point breaking like these composite lugs did.

[Chicago567]

Accident Investigations

According to Crain's Chicago Business last week, United Airlines has asked Boeing and the makers of Airbus for competing bids for up to 150 planes, a sales contract worth US $10-$12 billion dollars. Whether AF 447 went down because of a relatively inexpensive Pitot problem v. a massive problem with the tail fin/rudder design or something equally challenging may impact United's choice.

I used to work for the NHTSA investigating fatal truck and car accidents. The process for aircraft is the same: one uses the scientific method and sets up competing hypotheses that explain ALL of the existing scenarios, and then rigorously DISPROVES these hypotheses one by one with ALL existing factual data (even the LAV message will be investigated). The last hypothesis that cannot be disproved is most likely to be correct. The Pentagon already followed this process, presumably based on the ACARS messages, when they said a bomb did not bring down the plane.

I wonder why the Pitot icing is already the most likely explanation with so few data. The NTSB, I believe, would have raised the issue, but would have gathered more data before they disproved all other competing hypotheses. In my experience with mechanical systems and autopses, the evidence from the a/c remains and bodies will be equally valuable as the FDR (which may have been filtered) and the VDR.

[Mad (Flt) Scientist]

Originally Posted by gem1

This may provide a clue as to what broke first.

No it won't. Because it seems most likely that the TCAS was reporting a fault due to erroneous airspeed or altitude data - just like half the systems on the aircraft at that point. (Exaggeration, but you get the idea)

It's already been pointed out that 23 of 24 (at least) ACARS messages share a potential common cause in air data sources. It would be dangerous to assume an independent cause for one of those messages at this point.

[edmundronald]

Any plane can be shaken apart

Here is a nice link about the AA587 crash details and the controversy about composites. As an engineer, I like the writing style.

[Mad (Flt) Scientist]

Would one of the many people taking the idea of a TCAS Antenna fault and running with that idea please provide validation of that interpretation of the reported ACARS line referring to TCAS, i.e.

34 43/06 WRN WN0906010210 344300506 NAV TCAS FAULT 09-06-01 AF 447

Since I've seen a number of statements that this is not an antenna specific fault, and is entirely consistent with a faulty/erroneous/unreliable air data input. (Including one reference to a QRH procedure)

[bobrun]

Wasn't there more failures than those related to the air data?

Loss of pitot tubes or ADRs wouldn't explained the failures of PRIM1 and SEC1 flight control computers. It wouldn't cause the aircraft to lose pressure either.

Also, the failures were reported by the CMS computer over 2-3 mins, but that doesn't mean they happened over that period of time.

[Mad (Flt) Scientist]

Originally Posted by bobrun

Wasn't there more failures than those related to the air data?

Loss of pitot tubes or ADRs wouldn't explained the failures of PRIM1 and SEC1 flight control computers. It wouldn't cause the aircraft to lose pressure either.

Also, the failures were reported by the CMS computer over 2-3 mins, but that doesn't mean they happened over that period of time.

The search function is a wonderful thing. link

Originally Posted by captainflame

From a pilot prospective and arguably by just looking at the failures on the list:

2- One ADR failure is NOT consistant with flags on PFDs. (airbus eliminates 1 erroneous data source if 2 others are the same, transparently to the pilots)
BUT
Two ADR failures is ! and the ECAM message would read:

3- NAV ADR DISAGREE ! which we find in the list.

4- which in turns get the Airbus to revert to ALT LAW without protections (overspeed, and alpha)

5- autopilots, ATHR, NAV TCAS and F/CTL RUDDER TRVL LIM are associated inop items.

6- F/CTL computers (Prim 1 and SEC 1) could then be reset, and recovered (or not) (they receive wrong data from ADRs)

We're left with the STBY probe, feeding directly (no ADR) the ISIS (self contained stby instrument). Which also seems to report faulty info !!

Near Max rec ceiling, the operational airspeed margin on a jet is quite narrow. easy to go to overspeed, or a stall (no protection), especially in very turbulent conditions.
Flight manually assured using good old techniques of "Pitch and Power", in the QRH for "unreliable airspeed"

7- last message on list : ADVISORY pertaining to Air cond/pressurisation MODE Fault, seems to be linked to cab descent rate (unable to cope with a high VS descent to maintain required DIFF pressure !)

The F/CTL computers will report faults if they get enough bad air data info. The only message which doesn't seem to be directly air data driven is the last one - which may be a result of the aircraft having at that stage departed controlled flight. Or could yet be a strange air data driven one, if the cabin pressure controller is being confused by very strange air data.

And someone familiar with ACARS has mentioned that it takes the system fractions of a second to process a message and transmit it, so we can be reasonably sure that the timings are ROUGHLY correct, at least in terms of the to-the-miute timings.

[ChrisVJ]

I declare my amateur status, just another seeker after information.

I would think, in view of the importance of Pitot tubes, that they should be at least of different types so that a common failure did not cause them all to fail at the same time.

Are all the computers packed in the same bay/compartment? If so does not sound too redundant. (I do understand the electronically redundant thing, but physically safe? Chance in a million means sooner or later it will happens, at least that is my philosophy.)

That distance from land would they typically be using VHF or HF? (re HF aerial in vert. stabiliser.)

I understand three pitot tubes and three ADIRUs. Do A330s also have a steam ASI? If not I can imagine that all three pitot failure would be very difficult at high altitude, turbulence, even moderate, and in the dark.

Is/are AoA indicators part of the pitot tubes?

Are the Pitots electronic, ie, are there sensors in the pitot heads or are there still air pressure tubes back to the ADIRUs. Are the static heads duplicated for each pitot head, do alternate static cut in automatically, by electric switch or (like mine) by pulling a plug/ turning an air cock? (Or am I being naive assuming there is alternate static?)

[Willoz269]

LHP

I think you missed the whole point of the story....it is NOT the autopilot, it is the weather phenomenon that we were trying to highlight!

This guy nearly lost the aeroplane on manual, and the Boeing autopilot would have, in his opinion, led to a crash had he not disengaged. He counts himself as very fortunate to have come out of it alive. He questions if AF447 found a similar or more severe upset and didnt make it out alive.

[Captain-Crunch]

Christy Christy Christy,

Wikipedia is not a source either. If you re-read my post slowly again, you will find that I said it is a summary, and that it is constantly changing.

The ACARS list you keep putting up has errors in it. It is an older version of the new list updated by poster "selfin", a regular poster here who was kind enough to take the time and hand-type the leg report. He has updated it several times. We have communicated by email. If you want to put something up, please go to the trouble to get the lastest information. It is here:

http://www.pprune.org/4975386-post42.html

Now I'm sure you're a great pilot in your own right. But that doesn't make you an expert in Accident investigation. I witnessed your vindictive aggressiveness on the 123-Go Airlines thread, and I must say I thought it was detrimental to air safety as all you accomplished over there was to start a which hunt and put the poor slob out of business (as opposed to working with him to fix the problems.) You appear to have a sensationalist agenda. Most of the posters here are not out to smear Airbus or Air France. We understand that Aviation is a challenging endeavor for both employees and owners. We are just trying to get as many EDUCATED inputs as possible.

My hat is off, however, to the French Pilot's Union for taking a stand on defective pitot tubes and refusing to fly.



CC

[Captain-Crunch]

Kulwin Park,

It's possible the airplane didn't hit the water intact. But so far, from the scraps of clues we have (lack of a red cab alt warn, tight debris dispersal pattern if that is the case), tends to discount the likelihood that it came apart at hight altitude.

Because of the ACARS transmission requirements and characteristics and need for normal ships power, I feel it's safe to say that the airframe was intact at the time of the last transmission 0214z. Whether that Amber ECAM Advisory was a Cab decent advisory, or, an amber "Excessive Cabin V/S" has not been answered yet completely to my satisfaction by the techs here and other places in an affirmative fashion. They suspect that that's what it is; but the jury is still out even when they are referencing the 330 TSM and AMM's. It's not a red warning however imho.

Again, I don't see any acars evidence that a complete loss of cabin pressure occured prior to 0214z. (Maybe later in the dive at breakup.) But there is evidence of a severe loss of full flight control protections at 0213z caused most likely by loss of Air Data Computer function caused by loss of pitot air sensor data a few minutes earlier. (caused by, if previous airbus accidents reports are any guide: probe icing.)

Disclaimer: I have not flown the A330 and am unfamiliar with FBW.
But considering their wx and weight, a loss of control (jet upset) seems likely. Once you dive tens of thousands of feet in the dark like Adam Air did, high G's in the pull up are going to guarantee that things start coming off the airplane, (Cowlings and Fairing are the first to leave: we had two crews do this) likely followed by major structural damage of the forward and aft main spars in the wings and horizontal stabilizer. If you lose the rudder, then you're in bad trouble. But if you lose the vertical stabilizer (as was floating in the water), the game is completely over. I have examined airliners (as just a line pilot) after evasive flight control deflections, and the counter weights come off the control horns of elevators and large sections at the tips break off (if the stab is designed properly).

Some flyers brought the subject back up of Dutch roll at altitude. Unfortunately, airbus training that I've been in, doesn't demonstrate how to counter the effects of dutch roll if all yaw damper function is lost (the: that's never going to happen mentality). If line pilots try to use the rudder to right the aircraft, they will loose control even if all the flight controls are working properly. The correct way to keep flying at altitude is so tricky with the roll spoilers only (opposing every roll motion AFTER the roll is at it's zenith), that you are strongly advised to slow down and descend to a lower altitude immediately, like say FL290. This is sadly, going to put you is the bad weather were you don't want to be. But if you don't do it, I'm afraid not even old Mr. Neil Armstrong could save this one. Maybe at Edwards on a nice day, but not in the dark, in a storm with failing flight controls and failing instruments at the same time.

If all yaw damp function was lost, this would have quickly become a compound emergency that 90% of line pilots could not control even if they were "sticks".

My experience with rudder limiters is that it is expected that it will fail in its present mode. For example, if it fails at FL350, in high speed/low travel mode, you will loose rudder travel capability at low/slow altitude. If it fails in low speed mode, and now you're indicating fast, you could break the tail off with even normal rudder inputs. The problem is, with AF447, nobody knows at 350 in cruise what mode it failed in with the airspeed all over the place.

Right?

CC

The above, as all my posts are, are just my opinions only.

[Mad (Flt) Scientist]

Originally Posted by ChrisVJ

I declare my amateur status, just another seeker after information.

I would think, in view of the importance of Pitot tubes, that they should be at least of different types so that a common failure did not cause them all to fail at the same time.

Dissimilar components cause as many problems though. If you want to rely on being able to compare the outputs, then the components have to be practically identical. In particular, you're forced to design to at least the same specification, and perhaps the actual same design details (in the case of a pitot probe, geometry for example). So the same extreme conditions which will disable probe A will quite likely also disable probes B and c anyway. meanwhile you have to deal with the nuisance of 2 or 3 suppliers for one part, carrying more spare parts in the maintenance pool, and so on.

This may well not be common failure - in the sense of the probes being at fault - it may be the case that the environment was too much for them.

Suppose you wish to drive your car quickly, and put high speed-rated tyres on all 4 wheels, but to "be safe" you go for one Michelin, one Goodyear, one Bridgestone and one (someone else). Then you drive at 20mph over the rated limit for all the tyres anyway. The fact that they are dissimilar sourced won't do much good, in all likelihood. You're pushing them all outside their design spec anyway.

I understand three pitot tubes and three ADIRUs. Do A330s also have a steam ASI? If not I can imagine that all three pitot failure would be very difficult at high altitude, turbulence, even moderate, and in the dark.

ISIS is the "integrated standby" for when the main three systems fail - but it still has to source pressure from outside, so again, the environment could well be the issue. Not the aircraft components.

Is/are AoA indicators part of the pitot tubes?

No, different component. Some a/c use measured AoA to correct the measured air data - I don't know if this is the case for A330 or not.

Are the Pitots electronic, ie, are there sensors in the pitot heads or are there still air pressure tubes back to the ADIRUs. Are the static heads duplicated for each pitot head, do alternate static cut in automatically, by electric switch or (like mine) by pulling a plug/ turning an air cock? (Or am I being naive assuming there is alternate static?)

Air pressure lines to the ADIRUs, I believe. Most large a/c will have duplicate (left/right) statics for each pitot, to balance sideslip effects; again, I expect that to be the case here. The back-ups are through the multiple ADIRUs and ISIS - I doubt the crew can reconfigure the system beyond selecting which ADIRU they can display data from.

I'll remove my speculative statements if/when someone more knowledgable comes along

[Captain-Crunch]

ChristySweet:

You then wrote; 1 ) asking for any proof of a " turbulence encountered " message, again the link I provide has the list of received messages -and sure enough , there is NOT a" turbulance encountered" message. ( Voice, ACARS or keyboard..)

I don't think so. A message about turbulence is NOT going to be sent on that list to maintenance. Don't you understand this? Those are all AUTOMATED messages sent within miliseconds of the computer detecting the comparator disparities to the MECHANICS.

A human hand on the keyboard is likely to send a turbulence report to????????
Flight Operations! Assuming he pushed the right ACARS MESSAGE DEST button in turb!

CC

[philpop]

CC,

In this scenario, how would you explain that radio was not used for MAYDAY