Go Back  PPRuNe Forums > Flight Deck Forums > Tech Log
Reload this Page >

Air France A330 accident

Wikiposts
Search
Tech Log The very best in practical technical discussion on the web

Air France A330 accident

Thread Tools
 
Search this Thread
 
Old 30th Jun 2009, 12:03
  #121 (permalink)  
 
Join Date: Nov 2001
Location: Somewhere in the Tropics UTC+7 to 9
Posts: 450
Likes: 0
Received 0 Likes on 0 Posts
I have noticed that two pilots who regularly post sane comment about this issue have done a lot of work to try and figure out what may have happened.

One poster (Mandala499) who is probably the most sane guy there, has done a good job of trying to figure out the timeline of the ACARS messages and re-order them according to their respective time stamps.
Me, (aka Mandala499) am not a pilot (although some have refused to believe it *sigh*) . I just have a "deep interest" in the accident... but hundreds of pages of printouts etc, and that list, amongst with my other partner in crime there, still has a LOOOONG way to go...

I didn't get involved in the other AF447 topic because despite the heavy moderation, it appeared to be going in circles most of the time... and missed the one here.

That list still needs revision... it's been revised hundreds of times already.

I do need help on those just to see what happened based on the ACARS (to the limit of what the ACARS can give). How I think (personal theory) the plane went up to 0210 is on the works too but I keep coming back to the ACARS... although I need to match it up with the last position and where it was in terms of the weather (which sadly, indicates they made it (or almost) out of the weather, only to be hit by something such as icing). This doesn't rule out the occurence of intermitten errors/failures which would not have been picked out by the CMC... but that would be guessing too much. So pardon the inaccuracy if you find one, and help improve it.

I still need to know the details of 2CE2, and 22FN-10FC... (the part coding convention does give clues, but not much)
I find it interesting that Pitot 1 PHC came up, but ADR2 data downstream of the PHC2, failed.

As to difference between F/CTL ADR DISAGREE and NAV ADR DISAGREE??? No answer for the moment... it's equally baffling. However, it seems that the FCOMs I read have NAV ADR DISAGREE but not F/CTL DISAGREE. It might have something to do on where/how the disgreement happened... I was hoping that you guys over here can help.
PK-KAR is offline  
Old 30th Jun 2009, 12:18
  #122 (permalink)  
 
Join Date: Aug 2002
Location: Switzerland, Singapore
Posts: 1,309
Received 0 Likes on 0 Posts
My guess is that F/CTL ADR DISAGREE is showing when the FCCs have problems with ADR data, while NAV ADR DISAGREE comes on when the FMGC has problems with it.

You have to understand, that those modern planes take their data from all kind of sources and route them to all different aircraft subsystems. FCC needs the ADR data for computation of stearing commands while the FMGC needs it to compute navigational data.

Makes sense?

Anyway, as mentioned elsewhere, I do not think that we will get important information about the cause of the crash from the well known ACARS signals.

hth,
Dani
Dani is offline  
Old 30th Jun 2009, 13:43
  #123 (permalink)  
 
Join Date: Jun 2009
Location: Earth
Posts: 79
Likes: 0
Received 0 Likes on 0 Posts
To PK-KAR

Nice to 'meet' you PK

I will try to elaborate some constructive criticism. I am aware that it might bump into some serious hard work, please take it as I intend it to be, positive thinking to improve general knowledge.

Starting from your work, which I take is based upon the famous maintenance messages, two things strike me :

is the 34-11-15 fault to be taken as all 3 probes failed together, or as n°1 heating failed ? Finally, how many probes are deemed failed as we can deduce from this message ?

The devil hides in the details :

in the second message, 0210Z 27-93-34-0, I would like your thoughts on the tiny detail here : ADIRU1 BUS ADR1-2 TO FCPC2
Is the translation "ADR 1 & 2 no signal to FCPC2" correct ? Could it be instead that this BUS ADR1-2 TO FCPC2 refers to the part that connects ADR1 to n°2 parts of the system, here FCPC2 ? Wouldn't a dual ADR failure be named ADR1+2 instead ?

This hypothesis from the fact that ADRs are connected crossways to various parts of the system, and not only n°1 parts. And the fact that multiple failures are usually labelled with a + sign.

Best regards.
Svarin is offline  
Old 30th Jun 2009, 15:22
  #124 (permalink)  
 
Join Date: Jun 2009
Location: Earth
Posts: 79
Likes: 0
Received 0 Likes on 0 Posts
To Dani and PK-KAR,

please humor my insistence on this point : we need to seriously evaluate the difference between NAV ADR DISAGREE and F/CTL ADR DISAGREE.

From various documents I have read (and believe me, I have read quite a lot in a month, and I need to stop else I won't pass my medical because of eyesight), all instances of speed indication failures in the known incidents I could put my hands on were of the following nature : a quick determination of a F/CTL ADR DISAGREE. This failure is well documented, have a look at eurocockpit.

On the AF447, we have a slow-in-coming NAV ADR DISAGREE.

It might be minor, but I doubt it. AB shows pretty consistent wording throughout their 'namespace'. Since we are oriented towards a similarity between previous incidents and this dreadful crash, we might miss the exact difference which spelled death for so many people. I think we should try to find these differences, and surely some of them are inside the aircraft computations. This NAV / FCTL discrepancy definitely looks like one.

Of course, it 'very likely' means that the system which slaps the ADRs out is different in the instances. However, it also means that the overall reaction of various systems throughout the aircraft will be different on the AF447.

How different ?
Svarin is offline  
Old 30th Jun 2009, 17:08
  #125 (permalink)  
 
Join Date: Jun 2009
Location: Somewhere out there
Age: 39
Posts: 65
Likes: 0
Received 0 Likes on 0 Posts
ir2

I still need to know the details of 2CE2, and 22FN-10FC... (the part coding convention does give clues, but not much)
I find it interesting that Pitot 1 PHC came up, but ADR2 data downstream of the PHC2, failed.
How did you get to the relationship between air & phc data and the IR2 fault reported by disagreement with IR1 and IR3 ?

I think Inertial Data is independent from air data - and that the opposite is not true - so an IR fault would only be detected by comparison with inertial data...

Am I Wrong ?

P.S.: I know AOA differs from pitch attitude - it relates to the relatively motion of the airflow to the wing - but is there any inertial component in the AOA calculation ? This question is related to the qantas incident...
augustusjeremy is offline  
Old 30th Jun 2009, 18:09
  #126 (permalink)  
 
Join Date: Nov 2001
Location: Somewhere in the Tropics UTC+7 to 9
Posts: 450
Likes: 0
Received 0 Likes on 0 Posts
IR and ADR linkage?????

FCC needs the ADR data for computation of stearing commands while the FMGC needs it to compute navigational data.
My guess is that F/CTL ADR DISAGREE is showing when the FCCs have problems with ADR data, while NAV ADR DISAGREE comes on when the FMGC has problems with it.
I looked at 1 QRH from 1 airline, and 1 FCOM3 from a different airline... one has F/CTL ADR DISAGREE, the other has NAV ADR DISAGREE.

NAV ADR DISAGREE
This caution is triggered by the PRIMs, when they only use 2 ADRs, and these 2 ADRs disagree. This may occur when:
- One ADR has already been selected OFF by the pilot, or
- One ADR has been eliminated by the PRIM, without any caution, because it deviated from the others.
F/CTL ADR DISAGREE unfortunately has no descripton.
The procedures are identical (both versions are roughly from the same period)..

AIR SPD... XCHECK
- IF NO SPD DISAGREE
AOA Discrepancy

- IF SPD DISAGREE
ADR CHECK PROC...APPLY

F/CTL ALTN LAW (PROT LOST)
Note: Following an ADR disagree detected by the PRIMs, the ALTN LAW is latched. Resetting the PRIMs using the pushbutton will not allow recovery of normal law.

- MAX SPEED... 330/.82

STATUS
- MAX SPEED... 330/.82
CAT 3 SINGLE ONLY
RISK OF UNDUE STALL WARNING
RUD WITH CARE ABV 160 KT

INOP SYS
RUD TRV LIM
CAT 3 DUAL
Where one has NAV ADR DISAGREE, it doesn't have F/CTL ADR DISAGREE and vv.

Both are caused by the PRIMs rejecting 1 ADR and the other 2 disagree with each other... but are they identical? I guess this depends on the airline on where it wants the ADR DISAGREE to appear... coz, I looked at an Airbus document, and it just says "ADR DISAGREE"... *shaking head even more in confusion*. Perhaps it depends on the FWC standard one is using.





is the 34-11-15 fault to be taken as all 3 probes failed together, or as n°1 heating failed ? Finally, how many probes are deemed failed as we can deduce from this message ?
To be honest, I am not sure it means 3 of them failing together...
EFCS1, AFS - PROBE PITOT 1+2/2+3/1+3 (9DA)... AFS senses fluctuation or rejects an ADR, asks EFCS1 to compare the rest, 9DA as problem's suspect. I have a separate list showing 34-11-00, failures of inputs... most of the list is sensor power supply and switching... 34-11-15 is something else... (and different airlines may have different numbers).

ADIRU1 BUS ADR1-2 TO FCPC2 yes ADIRU1 BUS ADR1+2 TO FCPC2 makes more sense...
The thing is, I don't know the databus layout... ADR 1+2 Failure I think would occur if PRIM1 or none of the PRIMs receive the data... unless PRIM2 was the Master Prim at the time...
The "-" instead of the "+" could have just been a typo...


The devil hides in the details :

INDEED !!!!!

we need to seriously evaluate the difference between NAV ADR DISAGREE and F/CTL ADR DISAGREE.
Svarin, read the above... and I need new glasses now because of the same thing! Still no satisfactory answer though...

How did you get to the relationship between air & phc data and the IR2 fault reported by disagreement with IR1 and IR3 ?
I'm just reading what I got...

I think Inertial Data is independent from air data - and that the opposite is not true - so an IR fault would only be detected by comparison with inertial data...
That's what we all believe in... but somehow there's a relationship somewhere... be it just a crosscheck/diagnose function, or a relational function... The thing that caught me was that the FPV would disappear... FPV would largely be a result of all things IR and the only ADR part is barometric... but that's a downstream relationship... the other thing that caught my eye is the report from the VH-QPH upset (26Sep2006) and VH-QPG ADR/IR fault (27Dec08), where it too found a relationship between ADR and IR when one of them of the same ADIRU was faulty and/or switched off (unless my eyes deceive me)... But the wording from the manuals seems to say that they both perform their computations independently from another... But if they need data relationship, surely the loss of one function over the other should just mean degradation and not failure...

Still got a lot of reading to do (and repeat readings)...

PK-KAR
PK-KAR is offline  
Old 30th Jun 2009, 18:12
  #127 (permalink)  
 
Join Date: Nov 2001
Location: Somewhere in the Tropics UTC+7 to 9
Posts: 450
Likes: 0
Received 0 Likes on 0 Posts
I know AOA differs from pitch attitude - it relates to the relatively motion of the airflow to the wing - but is there any inertial component in the AOA calculation ? This question is related to the qantas incident...
If it has an AoA estimator, there could be a link... where a PRIM takes the speed from an ADR and computes a theoretical AoA, which is then compared to the voted AoA... to see if it fits... How it comes up with the theoretical AoA can come from several sources... but I don't have details of this yet... so this could be a wild guess...
PK-KAR is offline  
Old 30th Jun 2009, 18:13
  #128 (permalink)  
 
Join Date: Jan 2006
Location: Ireland
Posts: 43
Likes: 0
Received 0 Likes on 0 Posts
ACARS MEssages Partially Decoded

the data available (ACARS
messages) at this stage:
-Does not suggest any loss of electrical power
-Does not suggest loss of cockpit display units

The majority of the messages are linked with unreliable airspeed
measurement
• Some additional messages are also consistent with unreliable
airspeed measurement but require further analysis (hopefully we can all try piece together some info from this)
• Others resulted from crew action

02:10 UTC Maintenance message
“PROBE-PITOT 1+2/2+3/1+3(9DA)”

2 different sources of trigger:
The EFCS detected a speed decrease greater than 30kts in 1s on 2 or 3 airspeed measurements.
The AFS detected that each airspeed information compared to the 2 others showed at least transiently a difference greater than 20kts

The Above Message, as per its design, triggered the following messages:
“FCTL ALTN LAW”
“AUTO FLT REAC W/S DET FAULT”
“FLAG ON CAPT PFD SPD LIMIT”
and “FLAG ON F/O PFD SPD LIMIT”
-Indicates that characteristic speeds information
(VLS, VMAX….) were no more displayed on PFD
-Information displayed on both PFD speed scales
* current airspeed, selected airspeed
* SPD LIM flag on bottom of airspeed scales
“AUTO FLT AP OFF” - Auto disconnect of AP
“AUTO FTL ATHR OFF” - Again auto disconnect of ATHR as above
“FLAG ON CAPT PFD FD” and “FLAG ON F/O PFD FD” A red flag FD was displayed on both PFD (attitude area) to indicate that FD were disengaged but selected ON

At 2:10 UTC message “F/CTL RUD TRV LIM FAULT”
Indicates that the FCPC detected a speed variation higher than 50kts in 10s on 2 or 3 airspeed indications
Indicates that the rudder travel limit is locked at the position corresponding to the speed immediately before the detected speed variation


02:12 UTC “NAV ADR DISAGREE”
-The Flight Control System had already rejected one ADR AND
-Out of the 2 remaining ADR’s, the Flight Control System detected a discrepancy on one ADR parameter

Other maintenance messages and cockpit effects pointing at
ADR system:
02:10 UTC
–“FCPC2 (2CE) WRG: ADIRU 1 BUS ADR1-2 TO FCPC2”
–“MAINTENANCE STATUS EFCS1”
–“MAINTENANCE STATUS EFCS2”

02:11 UTC
–“ISIS (22FN-10FC) SPEED OR MACH FUNCTION”
–“ADIRU2 (1FP2)”

02:14 UTC
–“MAINTENANCE STATUS ADR2"

In Relation to other messages triggered:
02:13 UTC
“F/CTL PRIM1 FAULT” and “F/CTL SEC1 FAULT” - SHows PRIM 1 and SEC 1 were either both switched to off or Reset.
"FMGEC1 (1CA1)”
02:14 UTC
“ADVISORY CABIN VERTICAL SPEED” - This indicated Cab Vertical Speed was greather than plus or minus 1800ft/min


All pieced together from a prelim round up and also trying to shed some light on this
celtic mech is offline  
Old 30th Jun 2009, 18:21
  #129 (permalink)  
 
Join Date: Jun 2009
Location: Earth
Posts: 79
Likes: 0
Received 0 Likes on 0 Posts
ADIRU1 BUS ADR1-2 TO FCPC2 yes ADIRU1 BUS ADR1+2 TO FCPC2 makes more sense...
The thing is, I don't know the databus layout... ADR 1+2 Failure I think would occur if PRIM1 or none of the PRIMs receive the data... unless PRIM2 was the Master Prim at the time...
The "-" instead of the "+" could have just been a typo...
- or +
This reasonably cannot be a typo, I have it as a scan of a computer-generated document. How did you go about translating the message into plain English ?
Svarin is offline  
Old 30th Jun 2009, 19:13
  #130 (permalink)  
 
Join Date: Jun 2009
Location: Earth
Posts: 79
Likes: 0
Received 0 Likes on 0 Posts
To PK-KAR

The thing that caught me was that the FPV would disappear... FPV would largely be a result of all things IR and the only ADR part is barometric...
In pure theory, there are at least two ways to come up with an FPV : either take pure inertial reference and point a vector using IR, or compute the -almost- same vector by subtracting AoA from pitch. The second method needs air data.

What you end up with is either a ground-related FPV (inertial) or an air-related FPV.

Which theory was used in the A330 ?

Well, answers, answers : from QRHs, on the B777 side we have an inertial FPV, which the QRH states is reliable and to be used in airspeed indications loss situation (unreliable airspeed ind.), on all A320/330/340 the QRH states FPV unavailable... Draw your own conclusions.
Svarin is offline  
Old 30th Jun 2009, 20:34
  #131 (permalink)  
 
Join Date: Nov 2001
Location: Somewhere in the Tropics UTC+7 to 9
Posts: 450
Likes: 0
Received 0 Likes on 0 Posts
Svarin,
You mean this?
http://www.eurocockpit.com/images/PFR447.png ???
From Eurocockpit - Archives
One last important thing that we must do : in order to let you compare the events experienced by the flight AF447 with those described in the technical note mentioned above, we would like to expose the actual list of the contents of the famous ACARS messages sent by the plane. In order to keep our sources secret, we have transformed the appearance of the original file, but the content has remained exactly the same.
That is why I thought a typo was possible.. how did they transform the original file? Some editing? Or? Hence the possibility. If you're the author or person who transformed it, then I have to take your word for it (but then, should have the answer)... but anyways...

How did you go about translating the message into plain English ?
The comments mentioned below the headers are just my take on what it could mean... They're not translations into plain English...

As to the FPV... yes, it needs air data on the 330... it is something I don't get as to if they can get the BUSS (works off pure IR), why not just make the FPV all inertial? Surely air pressure gradients aren't "that" steep? *that one was picked out of the air*...

PK-KAR
PK-KAR is offline  
Old 1st Jul 2009, 06:50
  #132 (permalink)  
 
Join Date: Jun 2009
Location: California
Posts: 30
Likes: 0
Received 0 Likes on 0 Posts
other aircraft types with pitot issues

icing:

FR Doc E8-20702


cracks:

Docket No. FAA-2009-0005; Directorate Identifier 2008-NM-164-AD]

"This proposal would require operators to inspect the pitot tubes. If cracks are found in the tubes, they would need to be replaced, per the instructions of CASA Service Bulletin SB-212-34-11, Rev. 1 (dated Feb. 27, 2008).

This proposed AD, which resulted from an MCAI originated by the European Aviation Safety Agency, was prompted by the discovery of cracked pitot tubes on Britten-Norman BN-2 aircraft, which use the same Thales Optronics pitot tubes as are found on CASA C-212s.

Cracked pitot tubes could result in incorrect readings on the pressure instrumentation (e.g., altimeters, vertical speed indicators and airspeed indicators), potentially leading to navigational errors. FAA estimates that this proposed AD would affect 32 aircraft on the U.S. Registry and cost U.S. operators a total of $2,560, or $80 per airplane. Comments on this proposal, which are due March 25, should be sent to the DOT Docket Web site:"

http://dms.dot.gov;


overheating:

"AIRLINE INDUSTRY INFORMATION

July 29, 2002
United Airlines aircraft aborts flight due to faulty airspeed indicator

A United Airlines aircraft flying from Los Angeles to Washington DC was forced to abort its flight after was smoke was detected in the cockpit.

The aircraft - an Airbus A320, carrying 122 passengers - had been in flight for about 40 minutes when smoke began entering the cockpit. The flight returned to Los Angeles International Airport, where the passengers safely disembarked from the aircraft.

According to officials, the smoke occurred after a Pitot tube (which indicates airspeed) mounted on the fuselage overheated. The aircraft was expected to return to service once the faulty part was replaced, reported Reuters."

Last edited by Towhee; 1st Jul 2009 at 07:26.
Towhee is offline  
Old 1st Jul 2009, 19:39
  #133 (permalink)  
 
Join Date: Jun 2009
Location: Paname
Age: 55
Posts: 21
Likes: 0
Received 0 Likes on 0 Posts
A few more pitot incidents may be found in the article below:
Airbus crash: can a triple-redundant system give false readings? (Tony Collins's IT Projects Blog)
JuggleDan is offline  
Old 4th Jul 2009, 00:11
  #134 (permalink)  
 
Join Date: Sep 2001
Location: SCOTLAND
Posts: 80
Likes: 0
Received 0 Likes on 0 Posts
Hi
Cam any of you guys help me out here (susbseuent to my post on this topic being deleted from R&N ( I am not a pilot nor a journalist, just a curious sort of person)

The crashed aircraft was invoved in a taxying incident in 2006 during which it suffered wing damge described as "minimal." The other aircraft involved suffered "substantial" tail damage ( see []www.jadec.de/years/all2006.txt)

are there any international regs regarding post repair inspections for this specific type of incident or are they company-specific ?
Any help would be much appreciated

peter
PETERJ is offline  
Old 4th Jul 2009, 13:43
  #135 (permalink)  
 
Join Date: Jan 2008
Location: PARIS FRANCE
Age: 77
Posts: 53
Likes: 0
Received 0 Likes on 0 Posts
Celtic Mech, thank you for your very clear explanation
Coming a little late, and about the Nav ADR disagree message, the BEA (French investigation preliminary report) explains it thus:
Signification : ce message indique que les EFCS ont rejeté une ADR, puis
identifié une incohérence entre les deux ADR restantes sur l’un des paramètres
surveillés.
That is: this message tells that the EFCS have rejected one ADR, then has identified a discrepancy between the two remaining ADRs concerning one of the parameters.
For what it is worth...but as they work closely with Airbus I think we can trust them on that.
NARVAL is offline  
Old 5th Jul 2009, 06:12
  #136 (permalink)  
 
Join Date: Mar 2009
Location: Middle East
Age: 52
Posts: 214
Likes: 0
Received 0 Likes on 0 Posts
Type of failures

@Svarin
in the second message, 0210Z 27-93-34-0, I would like your thoughts on the tiny detail here : ADIRU1 BUS ADR1-2 TO FCPC2
Is the translation "ADR 1 & 2 no signal to FCPC2" correct ? Could it be instead that this BUS ADR1-2 TO FCPC2 refers to the part that connects ADR1 to n°2 parts of the system, here FCPC2 ? Wouldn't a dual ADR failure be named ADR1+2 instead ?
Only a class 1 failure is triggering a cockpit effect. Class 2 failures are not indicated during flight and have no operational impact.

Quote TSM :

Faults triggering an ECAM MAINTENANCE STATUS

These faults have no consequence on the system operating conditions. They are always GO without any restriction.

Anyhow yr translation is wrong. FCPC2 detected a malfunction in it's crosstalk between commanding part and self monitoring part, regarding ADR 2 inputs. But this type of failure will not cause a rejection of the FCPC.

Quote TSM:

Loss of the ADIRU1 ADR Bus 2 Input in the FCPC
1.
Possible Causes

-
FCPC-2 (2CE2)



-
FCPC-3 (2CE3)



-
wiring of the DGI 10 Bus from the related FCPC COM side to the first terminal block



-
wiring of the DGI 10 Bus from the related FCPC MON side to the first terminal block






Hope to shed some light
h3dxb is offline  
Old 5th Jul 2009, 07:15
  #137 (permalink)  
kijangnim
Guest
 
Posts: n/a
Greetings
SVARIN FPV uses Ground Speed over Rate of Descent, both data are inertial (earth related)
 
Old 8th Jan 2010, 19:11
  #138 (permalink)  
 
Join Date: Sep 2009
Location: S. LONDONDERRY, VT
Age: 74
Posts: 1
Likes: 0
Received 0 Likes on 0 Posts
Rain And Sound At -51c

I have been studying the ice particle icing issue and would like to communicate with the pilot and others regarding the conditions experienced, especially the sound characteristic reported during the apparent rain event at -51C. He reports a "sound of the plane getting pelted like an aluminum garage door". It would be interesting to characterize the sound quality to better identify the conditions. There are also a number of other questions that he or others may have observed that may be meaningful.

I believe I understand the "ice particle icing" issue and would like to wrap it up. If the pilot you quoted wishes anonymity, that is certainly alright. I would simply like to receive his observations. As a point of reference, I am a chemist and a pilot.
FAERGUS is offline  
Old 10th Jan 2010, 22:22
  #139 (permalink)  
 
Join Date: Nov 2001
Location: UK
Age: 70
Posts: 76
Likes: 0
Received 0 Likes on 0 Posts
Just diverging slightly, I haven’t spotted any discussion in the 3 threads running on this accident to possible icing/blockage of the static vents (they are all in close proximity to each-other) and the effect thereof.
I find it slightly bewildering why, when the altitude was obviously bleeding off and the wings were level the crew appears not to have initiated a nose down pitch for stall recovery. Or can we possibly conclude that the controls were compromised i.e. jammed horizontal stabaliser, preventing this.
Backoffice is offline  
Old 11th Jan 2010, 19:05
  #140 (permalink)  
 
Join Date: May 2000
Location: Glorious West Sussex
Age: 76
Posts: 1,020
Likes: 0
Received 0 Likes on 0 Posts
Backoffice
when the altitude was obviously bleeding off and the wings were level the crew appears not to have initiated a nose down pitch for stall recovery.
How do you know this? Have you found the DFDR?
TyroPicard is offline  


Contact Us - Archive - Advertising - Cookie Policy - Privacy Statement - Terms of Service

Copyright © 2024 MH Sub I, LLC dba Internet Brands. All rights reserved. Use of this site indicates your consent to the Terms of Use.