@ JuggleDan

I am not about to dispute your technical French definition of the meaning attributed by the BEA to the 02:10:34 final position, but I'd rather suggest the BEA has further data surrounding the basic position, e.g. flight level etc.. which they have deemed not relevant or desirable to be released in the Preliminary Report. In this respect, I submit that the AAIB and NTSB concurred with the wording used both in the prime report and the translated English version.

The 10 minute position reports (irrespective of interpretations around the accuracy associated with them) placed the a/c on the UN873 center line and the GS over the last 10 minutes (463KT) was only 1 knot slower than the earlier period. Without the FDR/CVR, 'when' the deviation occurred is purely speculative.

mm43

"- .1/FLR/FR0906010211 34123406IR2 1,EFCS1X,IR1,IR3,,,,ADIRU2 (1FP2),HARD"


From the description of FLR messages in the BEA report, 4 systems identified an IR2 fault during the 1 min correlation window that opened when the first system detected the fault.

IR2 got priority in the message construct because it detected an internal fault.

Because of the X after EFCS1 in the message, EFCS1 only detected a class 2 fault (crew awareness not required).

IR1 and IR3 also detected the IR2 class 1 fault (they were not faulty themselves).

The exact nature of the fault is unknown.

No other system reported an IR2 fault during the correlation window, since the remaining 3 identifier fields separated by comas are empty. If an IR2 fault was detected afer the end of the correlation window the CMC would not have it reported.

We cannot deduce from the message what system detected the fault first.

Incidentally, since IR1 and IR3 were able to detect an IR2 fault, it appears there is some cross talk between ADIRUs undocumented in the FCOM. A few details related to the QF072 accident also suggest this.

funfly;
Yes, thank you.

However, should there be no speed information at all due to these failures, (I have had one such failure on the 767), with a stall speed around 200kts for a 330 of around 200k kgs at 350, (that's a guess - my charts only go to FL200 - anyone have anything higher?), keeping 350 to 500kts on the GPS (depending of course on wind !), will provide a rough basis upon which to fine-tune the power and attitude and should keep us safely in the air until we can sort it out.

For those aircraft without GPS, the Flight With Unreliable Airspeed checklist, of which there are several memory items regarding pitch and power, will tide one over until better times. As the autopilot and autothrust disconnect, the key is to leave everything alone. So often, we feel we must "do" something, when the best action in some cases is to do nothing, for a short period of time. The airplane was flying fine before the loss so the current pitch and power would be a good starting point from which to fine tune things. The challenge in heavy turbulence, mountain wave or if one is caught inside a thunderstorm is to let the airplane ride vertically while maintaining pitch and power - too fast is far better than too slow- pitch is key.

Sure is easy to say it and type it in the comfort of one's home, isn't it?....

DJ77;
Thank you for the information. Quick question - is it the "x" that designates a Class 2 message? My AMM and the BEA report suggest it's an asterisk, or are we talking apples vs oranges? Tx.

GB
With your radar tilt and a whiskey compass they'd have had primary and secondary horizons :8 ....... Oh well.

On a different subject, I note the NTSB prelim on the NorthWest incident came and went without much in the way of comment.

http://www.ntsb.gov/ntsb/GenPDF.asp?id=DCA09IA064&rpt=p

A while back someone flamed the crew's statement that the cockpit temperature and humidity had suddenly risen although I've heard anecdotal tales over the years that reference this phenomenon in the tropics in the upper FL's, albeit in aircraft without the layered automation to make it a significant issue.

Rgds.

24V

SOP question
 

You are in an aircraft just currently in a really bad patch. (At about 0209, say.)

You have lost all three pitot tube inputs leading to all three IRU's declaring sick. You are following the manual and flying attitude and throttle. And you are significantly but not dramatically closer to landing in Brazil than any other place.

Does the pilot try to fly on to Dakar going through the rest of the bad patch essentially running ALT 2 all the way or does the pilot try to fly through the storm, again, to get to Brazil running ALT 2 all the way?

Either way would you want to get down to warmer air to get the pitot tube function back?

Is there a standard procedure or is it the pilot's call? Might they have elected for the devil they knew than the devil they didn't know?

JD-EE

Hi,

Are you insinuating the BEA board is not independent ?
BTW ... I don't too much worry about what is in this preliminary report .. but instead I worry about what is not in .. EG .. some recommandations as required by the ICAO.

Bye.

JD -
 

Yes. And the same holds in English: it is generally assumed that the final point is included if it has not been specified one way or the other. So those that want to lambast the BEA for 'lying about the deviation' have convention on their side in both languages.

But note that this is only convention, and possibly that paragraph was originally written by some math professor who is used to thinking in terms of excluded endpoints (mathematically it is often common to specify a range as including the initial value and NOT including the final value). So the person writing it might not realize that it is ambiguous.

I think more likely though, from the very sketchy coverage of almost everything in the report, that we should consider this to be a preliminary report, and that the BEA did not have all of its ducks in a row when it wrote it; indeed, they didn't even have a count of how many ducks they had to line up. There are other places in the report where there is bad grammar in both French and English, and the best explanation for the bad grammar was that the sentence was changed at the last minute, probably after the report had already been translated.

G.P.S.
 

From SLF newbie. Having read this whole thread twice and being impressed by all the thoughts and skills brought to bear on this terrible tragedy,it appears not to have reached any serious conclusion. However,reading up on GPS in Wikepedia in the 'Basic Concept of GPS' paragraph the article states "Some GPS Receivers may use Additional clues or assumptions (such as using the last known altitude,dead reckoning,inertial navigation,.......)to give a degraded position when fewer than four satellites are visible.(see notes) end of quote. Given that the plane might have been in trouble and descending in various attitudes at the last ACAR transmission,could this position be degraded,and if so does the loss of height vary the position calculated by the GPS unit./ Sorry for the lack of technical termanology,I am not a flier only a passenger.

IRUs
 

Apologies If I am wrong, but pitot tubes provide ADR data, not Inertial Reference Data.

Unless a very strange comparison algorithm validates Inertial Data from Air Data - and not just the opposed - blocked pitot tubes cannot generate IR faults by themselves.

There is no clear indication of loss of all IRUs in the AF447 case.

- where do you get this info from?

Regarding route decisions, I asked the same at #1926 but received no answer.

"rolled back to 60kts. The auto pilot and auto *
throttles disengaged. The Master Warning and Master Caution flashed, *
and the sounds of chirps and clicks letting us know these things *
were happening."

Sorry guys/gals it would not be ethical for me to post the e/mail, but the above is a quote from a very recent airspeed roll back on a 330. (Quoted previously on this thread)
So it would seem that airspeed failures at height are still happening. What I'd like to know is what Airbus are doing about it! In this case the failure messages were remarkedly similar to AF.

PJ2

CFR messages seem to be built using a limited character set that does not include an asterisk. Instead they use an "X" after the system acronym. Same messages in a PFR appear modified with a "*" before the acronym.

RE: PJ2 (#3572) : agreed also

RE: funfly (P.178 #3567)

Quote:
GPS cannot give you airspeed - it gives you ground speed. It's airspeed that makes aeroplanes fly.
KISS - there can be nothing more basic and therefore reliable for measuring airspeed than an air pressure monitor sticking out into the air itself (pitot tube).
/Unquote

Agreed entirely. However, I would like to add that there are other means of measuring airspeed. Before going into that, let me expose my understanding of the problem with pitot tubes. They need to be ventilated by drain holes to get rid of water or (melted) ice particles. The problem arises when the drain holes are blocked. Let me therefore explain two schemes which don't need drain holes.

As an ex-glider pilot I recall that sailplanes used to have a venturi probe. Whereas the pitot pressure is total pressure equal to static pressure plus dynamic pressure, the pressure measured with such a venturi is static MINUS dynamic pressure. In sailplanes that pressure is fed to a pneumatic rate-of-climb indicator (variometer) which then indicates variation of total energy, that is altitude corrected for airspeed changes. Since the air passes right through the venturi, it doesn't need drains.

Another scheme which I have come across was envisaged to replace the AoA vane (I don't recall why someone would want to do that, except that it had to do with icing and/or ice detectors). It consisted of a small cylinder projecting into the airstream with 2 pressure holes 120 degrees apart at the 1 o'clock and 5 o'clock positions (airflow coming from 9 o'clock). A servomotor would rotate the cylinder about its axis until the pressures at the 1 o'clock and 5 o'clock positions are equal, which means that the 9 o'clock position is the stagnation point indicating the direction of the local airflow, just as an AoA vane would do. I am writing this from distant memory so that details could well be different. The point is that one could measure a pressure that is representative of dynamic pressure on the surface of a suitably shaped body at a pressure hole where the airflow is flowing past it rather than into the hole.

However, just keep in mind that any scheme to replace pitots may well import new problems that are not easy to solve.

regards,
HN39

IRUs
 

RE: augustusjeremy (#3578)

Quote:
Unless a very strange comparison algorithm validates Inertial Data from Air Data
/Unquote

A functional diagram showing the relations between ADIRUs and PRIMs on this forum many pages ago showed an arrow within each of the three ADIRUs from ADR to IRU, indicating that the IRU receives data from the ADR. I've been puzzled by this ever since. Does anyone know why this is so?

regards,
HN39

The limitation is imposed by ACARS.

Alternate Law
 

Lurking SLF here again. Is my understanding correct in that when an aircraft such as the Airbus 3xx-series enters alternate, or direct, law that in so doing the aircraft is being "handed over" to the flying crew in a less than desirable state? I know I'm probably simplifying it but I think my question should be obvious in intent.
Thanks.

Quote:
Are you insinuating the BEA board is not independent ?
/unquote

Probably no more or no less independent than any other government entity anywhere else in the world. They all have an agenda.

Given that the recurring problem vis a vis Unreliable a/s is a suspect in AF447, how is the pitot problem continuing to carry weight? Each occurrence of anomalous readings means frozen pitots? I suggest there is more to the story besides rampant heated pitot failure?? Other types don't present the same Pitot issues. Pitot/airframe mis-match?? Power supply issues? It is getting mathematically difficult to sustain the three frozen pitot postulate.

There was an AD late last year that addressed Prim reset after engine failure test in sim. At failure, a/c pitched to zero, disallowing pilot input, a separate power supply was installed on separate bus bar, as the a/c was not pitch responsive while computer reset?

Will

handover in less than desirable state?
 

rgbrock1

I'm a software engineer rather than a pilot, so a s/w based answer.

On handover the pilot may be provided with a less functional plane than normal.

However, this is because the s/w has 'seen' events that [its designers feel] prevents it from reliably providing the 'lost' functionality.

So your wording of the question was very pertinent. Is it 'more desirable' to have full [but untrustworthy] functionality or a more limited [but trustworthy] functionality? AB have naturally gone for the latter.

Regards, Peter

IcePack
 

I dont suppose they were in the midst of negotiating some very impressive looking weather at the time?