Times Online (Calls to Ground A33/340) – Eurococopit (Pitots – 15 Year Old Saga) – Radio Chatter (ACARS Msgs)


According to several sources, including Folha Online: (English Translation)

Since the beginning of the search for the black box of Airbus, the French Navy have detected several beeps, but after further analysis, these tracks were discarded for not registering the technical parameters of the signals.

bn1

The BEA has the wrong blood to do it!  Le BEA a du mauvais sang à se faire !

HT: Blog: Politique, émois et moi

No political system can survive the truth (well … we would like to believe)

Aucun régime politique ne peut survivre au mensonge (enfin … on voudrait le croire)


theaustralian-logo-sml2

The below article is from The Australian (i.e. Times Online) by Charles Bremner | July 01, 2009

AIRBUS is expected to face calls to ground its worldwide fleet of long-range airliners tomorrow when French accident investigators issue their first account of what caused Air France Flight 447 to crash off Brazil on June 1.

It is believed that the accident bureau will report that faulty speed data and electronics were the main problem in the disaster, which killed 228 people.

The European Aviation Safety Agency (EASA) is likely to be asked why it had never taken action to remedy trouble that was well known with the Airbus 330 and 340 series. Nearly 1,000 of the aircraft are flying and until last month, no passenger had been killed in one.

“EASA has a legal and moral obligation to get to the bottom of this problem now,” said James Healy-Pratt of Stewarts Law in London. “If there is a defective system and the aircraft is unsafe then it should be grounded.”

No much here.  No creditable sources. Just talking with lawyers…


eurocockpitbanner-g

AF447 : a 15-year-old probe saga ?

The BEA announced a «factual report» would be presented on July 2nd at 3pm. EuroCockpit could publish a synthesis on this accident the same day, just before the BEA conference. No doubt that journalists will then have very pertinent questions to ask to “whoever-wishes-to-answer”.

We believe that these two presentations will be perfectly complementary : chances are the BEA will not compare ACARS messages issued by AF447 and those emitted last year by aircraft F-GNIH and F-GLZN, in incidents that seem similar, except on their effects.

In the meantime, and before proposing a logical chronology of the full ACARS messages that we had published on June 23rd 2009 (see below), we will now explain the history of these Air Data References problems on A330 and A340.

These problems probably started back in… 1994. While a A330 prototype crashed, on June the 30th, in Toulouse (pilot error, no need to ask), Airbus starts to write on the topic, and there will be further writings until 2008. Needless to say, we have analyzed the following documents. They seem authentic and do not contain any mentions that would prevent us from communicating them as part of an analysis related to the tragedy of flight AF447.

We therefore begin with the analysis of an OIT (Operator Information Telex) issued by Airbus.

OIT1994

On November 24th 1994, there is an OIT on the A330 and A340, which describes the existence of incidents that may have led to the loss of the auto-pilot system and the auto-thrust, the flight control goes into “alternate” mode in an icing environment. It could be due to the malfunction of Pitot tubes caused by ice crystals (ie: generally found at high altitude …). These events are accompanied by alarm messages strangely similar to those AF447 had sent by ACARS.

The problem seems serious enough to launch a TFU. A Technical Follow Up, which EuroCockpit was able to obtain a copy. This TFU was issued in December 1995 and its 5th revision was published in November 1996. This tells us that in icing conditions around the cumulonimbus of the inter-tropical convergence zone, small crystals of ice can sometimes disturb the Pitot tubes, and the Air Data Reference can be disrupted.

tfu96

http://www.eurocockpit.com/docs/TFU_34%5B1%5D.13.00.005_5.pdf

And guess what are the associated alarms?

Yes, the long litany of AF447 messages: autopilot disconnect, auto-thrust disconnect, flight controls alternate law, etc..

In July 2002, a new OIT was published. It explains that the Air Data Reference problems are related to Thales AA Pitot probes, that the AF447 was equipped with.

OIT2002

This time, we knew exactly what the sensors problems were : the drains. A small defect of these drains could prevent the proper functioning of the entire probe. An incorrect measurement of the total pressure could lead to a miscalculation of the speed. Amazingly, the BEA has spoken about this, and in those terms, at a previous press conference in early June 2009…

From 1997 to 2008, the manufacturer will maintain a SIL, “Service Information Letter”, which incorporates all the available knowledge on the Pitot probes and their impact on the various maintenance programs concerned.sil08

http://www.eurocockpit.com/docs/SIL_34_084.pdf

In January 2008, it is written that the BA model (the new probes) could solve the problems. It is “the solution” to these problems and “these new probes are now available.” It is also stated that all Airbus models are involved, including the A330… But it is also written that preventive maintenance procedures could help keep the old AA Pitot, even if we know they could cause problems. Probably based on this SIL, Air France wrote the 34-029 Technical Note that we have discovered.

In 2007, Airbus issues a SB (Service Bulletin) with references A330-34-3206. EuroCockpit has seen a copy of this document, which was amended on November 12th, 2008.

sb140907

For copyright reasons, we cannot publish the SB and its revision. However a well known U.S. website might have a copy (http://forfaiture.freeservers.com/343206R1.pdf). Both documents concern Thales Pitot probes and indicate:

This Service Bulletin proposes the replacement of the three pitot probes C16195AA PN (ends 9DA1, 9DA2 and 9DA3) by PN C16195BA new probes.

(…)

In addition, this new pitot probe introduces a new external protective layer to prevent corrosion.

(…)

Classification: Recommended
Timescale Accomplishment: Accomplishment of this Service Bulletin is recommended at the earliest opportunity where manpower and facilities are available.

Of course, these writings concern flight AF447, since the airplane serial number (660) is clearly indicated in the list of aircraft…

Since these documents were approved by EASA, it seems clear that nobody has seen the need to turn the SB in AD, as it was the case in 2001.

We now can have a better understanding why EASA states that there is “no evidence” that the probes are incriminated in the AF447 accident.

Our understanding gets clearer every day: the trail of Pitot probes failing in certain atmospheric conditions has been under investigation for 15 years. After having a closer look at Air France, Airbus and EASA managers communication, we understand why none of them want to take the blame.

This is probably the reason why they are all pointing their finger at the big cloud in the sky.

Other PDF links cited in above article:

http://www.eurocockpit.com/docs/TFU_34%5B1%5D.13.00.005_5.pdf

http://www.eurocockpit.com/docs/SIL_34_084.pdf


radio_imagesRadio Chatter/ MSG BOARDS / DISCUSSIONS

You never know what you may hear on the radio! Below are the most recent on-line  aviation chatter and message board discussions.  Please understand the below are only chatter/discussions among aviation professionals world-wide and should not be considered fact until all official information is released by Bureau d’Enquêtes et d’Disinformation (BEA).

———-

COMMENT/QUESTION:ADR1 feeds DMC1 which displays on PFD1,ADR2 feeds DMC2 which displays on PFD2,ADR3 is a standby, and can be selected to be displayed on PFD1 or PFD2 by the pilots.

The data output is fed to various systems, the display system does not reject… that’s my understanding… the rejections occur in the other systems… I do fear some think of it as a single stream from ADR to the end via displays, then computer A then computer B, etc…

COMMENT/QUESTION: Incorrect, the FMEGC does not control the RTL, it is done by the SECs.

Thank you… I misread the manual… The RTL part of FE module of FMGC interacts with SEC… not controls it… SEC has the controlling..  Smile

—————

COMMENT/QUESTION: Can you make some simpler exlpanation of the chain of events, whenever you are ready, for us who cannot understand fully all of the technical stuff?

Yes, what we need now is the pilot’s view through the interface.

Once again, we have to be very careful and in this respect, using the Air Caraibes F-OFDF incident, though tempting would be a mistake.

The maddening aspect of this accident is that we don’t know why the ADRs started providing faulty data to the different systems… Blaming it on icing is discarding most of the info we have (on the F-OFDF, they had a TAT probe icing…and even here…..).

So, we’re just allowed to give a picture of the time between 0210Z and 0214Z which is, to say the least, very confusing to the pilots.

What happened before is for the investigators with far better equipment than we’ve had, and what happened after is in the realm of speculative reasoning.

—————

COMMENT/QUESTION: As far as I know, the autothrust will have been told to maintain a given speed – not a speed in knots, but a ‘Mach. speed,’ calculated by reference to both the ASIs AND the ambient air pressure, as measured by the pitot tubes’ ‘cousins,’ the static ports…..which can also get blocked by ice or water……

Typical pseudo-technical nonsense : An ASI is an Air Speed Indicator. Doesn’t measure anything in an airliner. A “Mach speed” is an idiocy as Mach number is a ratio, by definition without dimension.

The autrothrust functions in two modes :

1/-Provides a set thrust, when speed is controlled by the pitch control : Climb, descent, idle…

2/-Controls the corrected airspeed or the Mach in Spd/Mach mode.

—————

COMMENT/QUESTION: Ok, who/what is going to determine the last valid CAS? If you have a disagree annunciated, select ALTN Air data, still have a disagree. Which ADRIU is correct? The reason there are multiple systems, A/P, A/THR, PRIM, RUD LIMIT, kicking out is UN-reliable Air Data. The A/C is saying ” I am not sure what the Air Data is, you figure it out”

ok, as stated above, if all three have different Airdata, when ALTN is selected, you would still have a disagree, now a 2-3 disagree instead of a 1-3 disagree. which AIDRU is the RUD LIM going to lock to?

Since you wrote “select ALTN Air Data”, I have to assume you mean a selection not automatic due to degradation. ALTN here is ALTN Law… there is “no” ALTN Air Data on the BusFBW architecture.

OK, the Air Data on the 330 is multi-tiered, if I may use the term…

Sources to ADIRU:

A330PitotStatic

Each ADIRU has its own monitoring function and can self reject it’s own data (alias it can commit suicide to prevent a genocide).

That then splits into 3 data streams:

1. Into the what air data the crew uses for their respctive displays…

330AirDataSwitch

If the ADR used goes nuts, that PFD will show the air data going nuts… There is no “ALTN”… Just select which one you want to use (leave your norm). This is, unrelated to what the FCS does for air data source.

2. Into what the system uses…

A330FCSElecRudder

ADIRU here isn’t like the 777, it has 3 ADRs which it uses all the time, unless one or more or all fails. It uses the average of the 3 ADRs as a base value.

ADR Rejections:

– If one goes bonkers and goes astray on ANY variable for enough time, it gets shot by the FCS and dies… this is the conformance/democracy measure.

– If any one of the ADRs does a sprinter on the CAS, the FCS will shot that ADR dead… This is the Prison Measure.

– If two of them goes nuts and tries to run faster than the moving average, then the FCS shoots those 2 preserving the remaining one (unless that one commits suicide). This is the anti-riot measure.

– If 1 is shot dead, and the remaining 2 decides to break their coalition, the FCS will ask you which one do you want to shoot at…

See previous chatter post for the different rejection thresholds.

3. The ADIRU also sends data to the AFS… the autopilot…

If the autopilot sees any one of them jumping up and down rapidly, the AFS tell the FCS to shoot and kill that ADR… if 2 of the ADRs get shot and dies, the AFS will itself go into a comma and disconnects… same goes for the autothrottles. (Anti genocide measures).

COMMENT/QUESTION:  So, what will the rudder travel do? The rudder travel will follow what the FCS uses for the speed… ie: average of 3, if one gets shot, average of 2, if the 2 can’t agree, it’ll wait for which one you choose.

Now, remember, the ADRs can commit suicide themselves too, by comparing static data (each ADR has 2 static sources and ADR 3 doesn’t do this self check)… and the FCS does check on them to check they’re alive and valid… AoA Estimation is one method…

All the checks done by the FCS is designed to pick up errors early enough to prevent genocide or fraticide between AFS…

So what do you do with the dead/rejected/erroneous ADR? What if it tries to come back as a zombie and affect the system? Well, follow the ECAM and QRH… which is… BURY THE DAMN THING! Switch it OFF.

Sorry for the crude analogy.

What happens with the rudder limiter? (ALTN Law = Damper goes to 4deg limit)…

3 ADR valid = It’ll use the mean.

2 ADR valid = It’ll use the mean.

2 ADR alive in disagreement = It’ll take the last valid mean as RUD TRV LIM will appear on ECAM when ADR DISAGREE comes up.

1 ADR valid = If ADR DISAGREE condition occurs see above. If somehow not, it will use that ADR’s speed.

If we see in the F-OFDF case… the ADRs jumped and tried to do a runner…

Once there’s an ADR DISAGREE or 2 ADR Faults, the aircraft entered alternate law, hence the yaw damper is limited to +/- 4degs… the limiter? Well, the F-OFDF case saw all 3 ADRs doing a runner and rejected them all, so it was locked at about 270KCAS value… the ADR DISAGREE was a result of 3 being rejected.

However, for F-GZCP, we just don’t know at what speed it was locked at… It depends on when the RUD TRV LIM appeared. If it occured at the same time as the A/P OFF and A/T off, then it can be deduced that it lost all valid speed (FCS killed all ADRs)… so probably around 270KCAS as well… but we don’t know that for sure…

COMMENT/QUESTION:  What moves on an Airbus FBW plane to tell the pilot what the airplane is doing? there isn’t any mechanical feedback, only artificial feel.

Mechanical feedback? Trim wheel???  Smile

There is a fundamental difference on the Airbus… which is the aircraft obeys the command of the pilot except when protection is being executed. The stick tells the plane what the pilot wants… not “what elevator and aileron displacement” the pilot wants.

Except for direct law:

The pitch is load factor demand… ie: no stick input = 1G (be it nose up or nose down). This is the fundamental difference. A feedback into the stick means when it moves, it’s not doing 1G… Which is not what the pilot asked for it to do… (That 1G target is one reason why some pilots don’t like the Bus in turbulence… too “stiff” in trying to maintain 1G according to some)

On the roll, it’s roll rate… again, no stick input means I want to maintain that bank angle…. stick feedback would mean it’s not doing it…

Now on both pitch and roll, a feedback movement into the stick would be another command? or a reaction? or a self feeding circle of doom?

The feedback is in the control surface sensors, which sends the control surface displacement back to the FCS… that’s where it closes the feedback loop. If the control surface displacement doesn’t agree with what the FCS wants… it’ll scream… So the moving yoke feedback exists in the FCS’s virtual domain. The FCS is your slave and your guide dog at the same time (just feed it electricity and it’ll be happy  Smile )

foto4


Related Links

U.S. Department of Transportation Federal Aviation Administration Advisory Circular

Subject: CERTIFICATION OF PART 23 AIRPLANES FOR FLIGHT IN ICING CONDITIONS (pdf)

Date: April 19, 2007 – Initiated by: ACE-100 – AC No: 23.1419-2D

—————-

Flight Instrument External Probes

EASA Requirements –Flight in Icing Conditions (pdf)

Eric Duvivier,EASA Certification Specialist

Flight in Icing -Seville, September 24-27, 2007

—————

In Service Events

• Underlying Safety Issue

• Modes of bad acquisition

• 25.1309 approach

• A/C Requirements

• Issues

• EASA Conditions

In Service Events

• A significant numbers of in-service events reported in Europe icing conditions or heavy rain conditions

• Most of the incident reports: airspeed fluctuation while in severe atmospheric conditions

• Also: Temporary total loss of airspeed indications

• A significant incident also happened with an instrumented test aircraft during a certification test Provide more detailed data. Analysis of the atmospheric conditions

• Icing Conditions at an unusual high altitude and at very low TemperatureConditions outside the environment of PART 25 Appendix C

A330_Flight_Deck_Cockpit


END

Advertisements