AF447

[HarryMann]

I was waiting for 'something to break', and needed the Boeing 'yoke' to hang on to, and I do wonder how you cope with a tiny sidestick? Can you avoid involuntary inputs while being tossed around?

I doubt it somehow, anyone tried driving a 4x4 off-road with a joystick? Impossible I'd say, a proper steering wheel stabilises inputs and allows the torso to give a 'sensible' reference frame...

[lear60fellow]

I just read in this Spanish forum ( Una hipótesis a tener en cuenta (esta sí)) a theory based on an actual situation that happened to an IBERIA flight on May 2001, the actual captain was flying B747 and retired flying A340, it´s quite interesting as it happened in the same area. I will translate the basics.

"Flying from Buenos Aires we overflew Rio de Janeiro and followed the same route that AF474 was flying when the accident happened. Crossing the ITCZ at FL370 with moderate to heavy turbulence in a 1-2 minutes period we experienced a sudden increase in air temperature, from -48ºC to -19ºC. The consecuences were that we passed flying with a margin of 10.000kgs to being 15.000kgs above maximum weight and out of limits at that temperature and Flight Level. Inmediately the aircraft experienced a stall situation with lots of buffering. I desconected the AP y and we descended 4.000ft flying in the coffin corner having at the same time high and low speed stalls. If I wouldn´t have disconnected the AP and flew it manually we would be now into the deep ocean. The AP tried to maintain at all times the altitud using engines and that was impossible and the consecuences would have been an abnormal position of the aircraft and spacial desorientation to recover it at night and into the clouds.

I would describe it as an enormous 40 miles radius of hot air on the begining of a developing storm ascending to high altitude. After 5 min everything became normal and we continued our flight.

AIRBUS criteria says that flying into moderate or heavy turbulence "YOU MUST NOT DISCONET THE AP" but the fenomenom is not described or even contemplated and if the crew of AF474 did not disconneted the AP probably they had a situation similar to the one I experienced. The theory is that ACARS messages were sent due to this situation:


” New information provided by sources within Air France suggests, that the ACARS messages of system failures started to arrive at 02:10Z indicating, that the autopilot had disengaged and the fly by wire system had changed to alternate law. Between 02:11Z and 02:13Z a flurry of messages regarding ADIRU and ISIS faults arrived, at 02:13Z PRIM 1 and SEC 1 faults were indicated, at 02:14Z the last message received was an advisory regarding cabin vertical speed. That sequence of messages could not be independently verified.”

On basics it means: the AP disconnects when gets to the limit but the aircraft is already on an abnormal position, then is reverts to alternative flight control, then is followed by ISIS and Interial failures, then PRIM1 + SEC1 computer failures and finally a sudden decreas in cabin pressure (when the aircraft is already divided in parts), multimple electronics failures as the aircraft is desintegrating.

All this happens between 2:10Z and 2:14Z

[avspook]

woodpecker

WRT the SAARU on the 777, this was a mandated device the FAA made them installl as they did not trust honywells 1 in 3 billion chance of an undetected ADIRU failure.
Its input is AVERAGED with the ADIRU .. if the SAARU had been on MEL the Malasian 777 would have attempted to fly backwards,

SAARU was a cheap device using teh same FOG type RLG's as the ISFD (Boeing for ISIS) consequently they failed a lot.
New SAARUS have the SAME RLG's as the ADRIUS now. & we are back to a 1 in 3 billion chance.. Guess Malaysian got lucky

so on that aircraft You had the SAARU the Adiru & the ISFD.

If it lost the probe data you would be in a similar scenario. Does the ICE Detector work well on the A330?

[DC-ATE]

While theoretically you have two choices with this incident, Pilot Error or Aircraft/Systems failure, there really is only one choice; that of computer failure for whatever reason.

That is to say, being as how it was an Airbus with computerized everything, some part of the electronic system failed for some reason, and the pilots, no matter how well trained, could not cope with the situation. This seems to be apparent (computer problem/s) with all of the ACARS messages that were sent. Had the aircraft simply broken up, no messages (or at least not the amount) could have been sent.

On the other hand, pilot error by flying into an area of severe weather could have put the aircraft in a position the computers simply could not handle. One way or the other, I think all will have to agree, like it or not; the computers played a significant role in this incident.

The question/problem now remains to figure out how to rectify the computer problems that have affected not only this make aircraft, but others as well in the past. It would seem that computers are "here to stay" in aviation, but as the pilot gets pushed further and further "out of the loop", it would appear that we have more and more problems.

The pilot needs to have TOTAL control of his aircraft if need be at ANY time, even if it means exceeding the design limits. Basic piloting skills have been lost due to computerized flight. This fact has shown itself on too many occasions in past accidents. If this means re-installing cables, so be it. The pilot also needs at ALL times, analog flight instruments (standby 'steam gauges' if you will) to find his way to a safe landing when all of his computers fail. What good will it do to have inflight data being transmitted to the ground continuously no matter where on Earth the aircraft is, if it will not reduce the chance of similar accidents? What will it tell us? Part of the automatic system failed.....duh.

The whole idea of all this "progress" was supposed to make aviation safer. I fail to see where that has been accomplished yet. All that has been accomplished is that manufactures have been able to produce aircraft cheaper and the ticket price has gone down.....along with pilot saleries. There has been needless loss of life in recent accidents due to either loss of basic flying skills driven by computers doing the job or by those same computers failing when most needed.

[Captain-Crunch]

ACARS mystery codes

Swedish Steve said,

What makes you think that all three AD computors gave up?
In all the messages there is only one referring to the AD.
ADR Disagree. This infers that there is a fault with one AD output. How do you get to all three?

p.s. glad you said AD computors. Many posters here refer to ADIRU failures but it is only the AD part of the ADIRU that has failed, not the IRU part.
Today 10:51

Fair Question Steve,

I'm not basing it on the entry you mention the ADR Disagree WARNING. Because that ECAM warning only means "one or more" has a problem, right? It could be one, It could be three, we can't tell from this warning.

But unless I'm mistaken, that warning came one minute later from FAULT 34123406-IR2 1, EFCS1X, IR1, IR3,
As also an A&P, I agree with Greenspinner, that the Warnings are caused first by the Faults (FLR)


Photo snapshot credit: Machaca & Mods


Here are the codes given to us by techs on this thread:

2131 Cabin Pressure Controller
3410 Environmental Conditions
3411 Pitot-Static System
3412 Outside Air Temp
3422 Directional Gyro & Ind.
3443 Doppler

(JASC Standard Codes)

Captain,
At this stage describe all the messages would be little bit useless, as what we need is the fault message from CMS related to these Wrn. In fact we need the full Current flight Report rather that the current leg report provided here.
Nevertheless, as I got the A330 AFR TSM in front of me, I do try to understand ,as you, what’s occurs during this flight, and despite my 17 years of experience (as line engineer) on this type of machine (A340/A330), I’ve still have difficulties to sort it out.
However for your perusal
228334 – FMGC
341234 – ADIRU
279334 – EFCS
341115 – PROBE – PITOT
That’s all the data, which could be helpful with what we have right now
The WRn messages are only 4 digits exploitable and the sixth one are mainly 00

Cheers

Greenspinner said:
You're right Fargo, and I carried out trouble shooting starting from the fault message (and not the Wrng which are the result of the fault) and guess what.
341115 lost of all pitots probe (to make it short) not duplicable faulf on Gnd
341234 lost of all ADIRU'S, not duplicable on Gnd, and digging a little bit further, lost of ADIRU pwr (to make it short also),
and all these events occurs within 1 minute....

Quote:
Ahh,

thanks Fargoo.

So, "15" is the pitot probe. Does this mean the pitot probe signal is lost to the nav computers?

Next question: What number is AOA probe?
On the small Airbus we operate normally Probe anti-ice failure (be it probe element, probe heat computer or wiring between the two) would bring up a message in that Chapter. Could quite easily be a different case on the A330 though.

34-11-16 is Static ports.
34-11-17 is Air Data Modules.
34-11-18 is TAT sensor.
34-11-19 is AOA sensor.

To decode the messages and their possible causes you need someone with access to the A330 Troubleshooting Manual (TSM).

Hope this helps a bit.

Fargoo

Furthermore, JAUH is right, ISIS it's 3422, but this kind of error exist sometime in the CMS.
However, starting again from this fault messasge,ISIS (22FN-10FC) SPEED OR MACH FUNCTION, with the red flag on the ISIS (I assume) it's lead me again to check the Stby Pitot Probe failure (9DA1, 9DA2, 9DA3)!
As the first fault recorded its also Pitot probe (341115), I would suspect severe icing (this only engage myself)

2131 Cabin Pressure Controller
3410 Environmental Conditions
3411 Pitot-Static System
3412 Outside Air Temp
3422 Directional Gyro & Ind.
3443 Doppler

(JASC Standard Codes)

Last edited by jauh : Today at 04:32. Reason: added even more stuff

ADR 1 controls the speed tape and altitude display on the Captain's PFD.
ADR 2 controls the First Officer's side.
Stand-by ADR 3 (mostly fed from the same probes as ISIS) can control the Captain or First Officer's side, if required, with some hard switching.

In case of 'ADR Disagree' message, compare airspeed indications on all three displays (the last one being ISIS). If all are the same, the warning is due to an AOA discrepancy.
If one speed display is off by more than 16 kts, switch the corresponding ADR off and switch the PFD display over to ADR 3 (if required).
If all speed displays are off by more than 16 kts and reliable ADR cannot be identified, switch ADR 1 and 2 off and fly pitch and N1 setting as per the 'Unreliable Speed Indication' paper checklist.

Sorry, I was out off sight for a while, and I’ve got planes falling down here, but rather “smooth as Silk…”


Nevertheless, here after a breakdown of all the ATA messages related to the Current Leg Report provided.


34-22-25 - INDICATOR - ISIS (INTEGRATED STANDBY INSTRUMENT SYSTEM)

34-43-00 - TRAFFIC AND TERRAIN COLLISION AVOIDANCE SYSTEM
34-12-00 - AIR DATA/INERTIAL REFERENCE SYSTEM (ADIRS) ((ADIRU & CDU))
34-10-00 - AIR DATA/INERTIAL REFERENCE SYSTEM (ADIRS)
27-90-00 - ELECTRICAL FLIGHT CONTROL SYSTEM (EFCS)
22-83-34 - FMGEC (FLIGHT MANAGEMENT, GUIDANCE AND ENVELOPE COMPUTER)
22-62-00 - FLIGHT ENVELOPE COMPUTATION
22-30-00 – AUTOTHRUST
27-23-00 - RUDDER AND PEDAL TRAVEL LIMITING ACTUATION
27-93-00 - FLIGHT CONTROL PRIMARY COMPUTER (FCPC)
34-11-15 - PROBE – PITOT
27-93-34 - FCPC (FLIGHT CONTROL PRIMARY COMPUTER)
21-31-00 - PRESSURE CONTROL AND MONITORING
27-91-00 - OPERATIONAL CONFIGURATION (F/Ctl Altn law)

for Grebllaw, 341200106 it’s ADIRU and last 2 digits 06 mean phase 06 (Cruise)
Trusting this will be of some interest
Cheers
-
-
-

But I could be wrong Steve. And forgive me guys for not giving proper credit for your previous research and posts. I'm not that organized!

Note: All my posts are strictly my opinion only.

.

[jr hartley]

BBc news website

Faulty speed meters
Mr Arslanian confirmed that the missing jet had had a problem calculating its speed, adding that it was a recurring problem on the A330s and that Airbus was undertaking a replacement programme.






"We have seen a certain number of these types of faults on the A330," Mr Arslanian said. "There is a programme of replacement, of improvement."
But he insisted the planes were safe in the meantime

Thats going to make the public feel secure!

[Green Guard]

Only FDR and VCR can solve the mistery, but...

Heated pitots getting iced could mean only one thing:
flying straight into the cell,

but at night during lightings, (radar misuse is excluded here),
that can happen only if pilot or pilots in cockpit
were asleep at that half an hour, (prior to coming to the cell).

And if heated pitots iced, why engines would not ?

Consequently gradual electric power loss would come!

"All asleep case" would not be first case.
That’s why FA have to come to "visit" cockpit often during sleeeepy hours.

No parts of aircraft found may suggest pilots ditched the aircraft.

[Swedish Steve]

But unless I'm mistaken, that warning came from fault 34123406-IR2 1, EFCS1X, IRU1, IRU3,

Captain Crunch, Must admit I have never seen present leg faults presented in this fashion, I am usually looking at a PFR.
But isn't this fault IR2 fault, as reported by EFCS1, IRU1 and IRU3?

[safetypee]

PJ2, MFS, et al, apologies for yet again pushing TAT, but I wish to understand it’s potential (or otherwise) to contribute to the ACARs indications.
Thus seeking your technical expertise; what systems might an erroneous TAT affect? Considering both a large change of TAT - all inputs, or just one system in error and thus the potential to detect the difference?
What systems (computations) use TAT (e.g. ADIRU, rudder limiter via Speed/Mach, etc)?
I note the reported TAT input to the cabin pressure system.

Possible thread relevance to unreliable speed; http://www.pprune.org/tech-log/37578...ml#post4963257
Note other references to erroneous TAT “… sudden increase in air temperature, from -48ºC to -19ºC …” (#315). This is probably similar to the phenomena reported in an Australian BAe146 serious engine icing incident, which subsequently was revealed to be an indication problem due to frozen TAT probe – no airframe icing, but severe engine/probe icing.

[matthewsjl]

It loks to me like the ACARS FLR messages have supplemental data fields (some of which are pobably truncated on the leaked printout).

Could somebody explain what (if any) the significance of an 'X' is in the supplemental data fields. As an example one of the failures is:

Code:

34111506EFCS2  1,EFCS1,......

compared to the following entry (line higher on the printout):

Code:

27933406EFCS1  X2,EFCS2X,......

Thanks, John.

[Captain-Crunch]

Captain Crunch, Must admit I have never seen present leg faults presented in this fashion, I am usually looking at a PFR.
But isn't this fault IR2 fault, as reported by EFCS1, IRU1 and IRU3?

Correction: They are not listed as IRU's, rather they are IR's..sorry: typed what I wanted to see there....

Steve, It appears to be an internal mtc leg report broadcast on French T.V.
I don't have the A330 TSM or AMM to reference. I'm just going by the research of the techs here at pprune. But the codes have been independently broken down by several airbus mechanics, and thought there are small differences in nomenclature (common) between versions, they match, which gives me a warm-fuzzy. Have all three IR inputs faulted on that line? update: [GreenSpinner thinks so after research, but his colleague below affirms your impression that it's just #2IR (U)] Stay tuned.

The section is 34 (Nav) and the sub-grouping is 12 (OAT sensors) and the item status is 34 (ADIRU). If we could persuade GreenSpinner to scan us a T/S flowchart that would be awsome. But that's asking a lot. I'm not sure we pilots would understand it anyway!

CC

Note: All my posts are strictly my opinion only.

[forget]

If we could persuade GreenSpinner to scan us a T/S flowchart that would be awesome.

I just hope that Greenspinner () has not been put off by Odies out of place remarks. I don't think Odie realised that GS has spent 17 years fixing Airbuses and deciphering ACARS.

[avspook]

Quote:
Swedish Steve
But isn't this fault IR2 fault, as reported by EFCS1, IRU1 and IRU3?


That is correct.. An example would be a loss of data from IR2 data to the reporting LRU's

[Safety Concerns]

After trying to make sense of the acars messages the past 24 hours there is a scenario/situation that would explain 23 of the 24 acars messages. I would like to share it in order for others to decide if it has any loopholes but I don't want to be kicked for speculation.

Any ideas or post it on another thread?

[ihg]

Most of the speculation on the reason for the AF447 disappearance is based on two things:

- satellite images
- ACARS messages

With respect to the last one: Many of you think, that theses messages can provide a "sequence of events".

But is this really possible?

ACARS has not been designed as a "real time monitoring system". So, as I have no detailed knowledge about the ACARS implementation on the A330, I would have some questions before "abusing" ACARS as this:

- What is the usual internal sampling rate of the ACARS system? How often does it read out the error memory of the respective systems?
- How often does it usually send messages? Does it send messages as soon as a new system error appears? What's the time lag between error occurence and error message sent via ACARS?
- Can one message contain multiple errors? Or will each error transmitted in a separate message?

To put it the other way round:
If e.g. 10 errors would appear simultaneously, how would ACARS deal with that?
Assuming that each error would be send in a separate message, how long does it take to send all these messages?
Is the time stamp on these messages, the time when the message was sent or the time when the respective error occured?

Thanks in advance,

ihg

[Safety Concerns]

there are issues with acars but the messages sent are time stamped by the system reporting the failure. As all systems are on the same time bus you can get an accurate picture.

What you can't account for is the sampling rate of the system but this generally applies to normal conditions i.e. some air conditioning sensors are monitored every 3 minutes. However a significant fault or warning is time stamped in real time.

That doesn't prevent the scenario however that there were other messages relating to earlier incidents that didn't make it to the transmitter for whatever reason

[rouelan]

I note that, this morning, BEA talked of 24 ACARS msgs sent within 4 mns.
When comparing with doc leaked, I also count 24 msg.
Which suggests that BEA is working on exactely same piece of info.
Then I cant reconciliate with the fact that they pointed out a problem air speed incoherence ?

[positive climb gear]

ADIRS/IR Switching

My background is the A320 series, but believe the A330 is similar! I have noticed that some of the older aircraft I fly gives you the ability to switch off EITHER the Air Data or Inertial Reference feed to the respective ADIRS.

However, our latest aircraft just has the one switch- is this particular series of the A330 similar? thus would switching of the ADR also depower the IR reference also?

[OVERTALK]

Explanation for the Airspeed Anomalies/Discrepancies

Lots of discussion about pitot tubes instantly icing up as an explanation for the airspeed disparities/anomalies? There's a much simpler (Occam's razor) explanation - if you go back to first principles.

” New information provided by sources within Air France suggests, that the ACARS messages of system failures started to arrive at 02:10Z indicating, that the autopilot had disengaged and the fly by wire system had changed to alternate law. Between 02:11Z and 02:13Z a flurry of messages regarding ADIRU and ISIS faults arrived, at 02:13Z PRIM 1 and SEC 1 faults were indicated, at 02:14Z the last message received was an advisory regarding cabin vertical speed. That sequence of messages could not be independently verified.”

On basics it means: the AP disconnects when gets to the limit but the aircraft is already in an unusual attitude, then it reverts to alternative flight control, then this is followed by ISIS and Inertial failures, then PRIM1 + SEC1 computer failures and finally a sudden decrease in cabin pressure (when the aircraft is already divided in parts), multiple electronics failures as the aircraft is disintegrating.

All this happens between 2:10Z and 2:14Z

Looking at the incident described in post 316, we can imagine what would have happened to AF447, upon entering such a "warm pool", if the autopilot had remained engaged - and the incident allowed to develop. Quite simply, wing and nose drop plus yaw (aka autorotation) at a high power setting as the autopilot and autothrust tried to maintain the flight level - despite, due to the warm pool, suddenly being overweight for the environmentals.

It's called coffin corner. Aerodynamic stall and Mach buffet boundaries come together to define the upper edge of the aircraft's flight envelope. Autorotation (aka spinning) at a high power setting is going to have some further complicating ramifications:

a. firstly, engine intakes are going to be blanked and, being already close to their surge margins at that height, at least one engine will stall/surge...... complicating the autorotation with asymmetric thrust (as per the Jefferson City crash of a Pinnacle CRJ200).

b. Autopilot would disconnect and control law would change

c. ACARS transmitted indicated airspeed anomalies between left and right ASI's? In a spin, the F/O's pitot and static ports being on the other side of the airplane (to the captain's), the pitot and static sensed pressures will be in error (and also quite unalike) and so the ADIRU's are going to be initially integrating different air data and therefore calculating different speeds. That ADIRU disagreement will generate faults.

d. The situation at C will be beyond anything the ISIS has ever been programmed to cope with, so a fault message will be generated.

And thus the flow of ACARS messages begins:

Fourteen of those messages, says the Bureau d'Enquetes et d'Analyses, were sent within the space of one minute, from 02:10UTC to 02:11UTC.

This would have been immediately after the spin entry.

Subsequent ACARS messages (until 2:14Z) resulted from the descent into the storm clouds below and the transition into the unstable steepening spiral (G increasing until breakup thresholds were exceeded). The entire process (to breakup/no further messages) took 4 mins.

I'd imagine that, once into a spin, the centripetal(wrong) centrifugal forces on the flight-deck would've precluded any useful action or allowed any thoughts of transmitting a Mayday. The final cabin pressurization warning would've been immediately prior to total electrical power-loss (due to the first hull rupture preceding a general break-up due to the classic high-speed/g-tightening spiral).

Any amplification of this as a probable scenario (from the available ACARS messages)?

Any exceptions as to the likelihood of this exact scenario?

At least it would seem to credibly explain the airspeed discrepancies that otherwise seem to have everybody flummoxed. Not sure if the Airbus simulators could replicate this, but then again we've all seen what happens to airspeed in different types of spins (stable/unstable).... it fluctuates. Once the A330 descended into thicker air the spin would become an unstable spiral with its characteristics quite dependent upon residual thrust, C of G position and flight control trim settings. In the turbulence and in IMC with no functional flight instruments, a destructive break-up would be inevitable.
.
edit to change to "centrifugal"

[marsipulami]

Overtalk, you are drawing some very sensible conclusions regarding to that weather phenonoma described in post 316. All of a sudden you find yourself outside your envelope with consequently odd indications in terms of speed which might have led to the stream of ACARS messages. Though, we simply don't know yet.

I've heard a few similar stories of pilots crossing the ITCZ and encountering such a "pool of warm air" with the associated turbulence.
For them it was only a matter of time when they came out of the pool unharmed due their relative low weight at that time.

Marsipulami.