mm43

That is the bit that seems to have got them totally confused.

The AoA vane plus IAS provides the stall warning. Only problem is that the IAS must be greater than 60KTS for the combo to provide the SW. The SW stopped because the IAS had dropped too low, and when they attempted nose-down and got an IAS over 60KTS, the SW started again.

Their logic told them it was the wrong thing to do, hence the persistence with nose-up on the side-stick. I wonder if the actual AoA had been available, what they would have made of it?

beachbunny

It's reasonably simple, actually.

The stall warning system would only work within a specific range of parameters.
( An educated guess here)

The aircraft went though the range, to the "other side", ie excessive AoA, low fwd speed. When the pilot tried applying nose down, the a/c would have re-entered the operative range, hence the stall warning sounding again. Proves he was moving in the right direction, be it elevator input, or THS, (except that it sounds like the THS had not started to wind back the other way, before the ND was discontinued)
Had the ND input continued, the rollover would have continued, (with the stall warning still sounding,) until AoA reached or passed through, the "onset" parameter. We now have control.

mm43

That's the next problem! Once the IAS has dropped below 60KTS, the aircraft is in Abnormal Law and the THS must be adjusted manually.

takata

Or that their AoA exceeded 30 deg (same effect = abnormal Law).
Reduce AoA, airspeed will increase, she will return to Alternate Law and the auto-trim will work again.

matthewsjl

My read of the above is that the plane was flying fine when the autopilot/thrust kicked out due to unrelaible airspeed. No power changes are mentioned by the BEA so the thrust required for level flight at FL350 and 0.8 must have been set and being produced. The initial control input is nose up (and left roll) and a climb begins. The plane isn't stalled at this point as a) it's capable of a climb and b) the BEA confirms that the last 3m30s were in a stalled condition which puts the start of the stall at about 2h11m (the stall warn triggers again at 2h10m51s). There are some nose-down inputs mentioned at this point and it seemingly controls the climb but is still probably in a climb (and bleeding airspeed) due to the pitch and VS indicating +700ft/min.

It appears that there was a trade of speed for altitude but no actual stall until 2h11m.

Once the stall begins, all BEA references are to nose-up up inputs and the THS rolls to full nose-up. Some references to nose-down are made later but as others have commented, with full nose-up THS it's a good question how recovery would be made at that point.

Shadoko

Hi,

Is this "Abnormal Law" dispatched for the PF?
Doesn't be curious BEA note doesn't not mention this?

One more question:
Why ISIS speed display goes down? Isn'it inertial? I don't understand how they can truly loose all that speed before ascending.

JD-EE

mm43, if it got them totally confused it's a software specification design problem. If the airspeed is less than 60kts, suddenly and the plane is 10k feet above ground for the plane's estimated location it's "obviously" stalled. It cannot fly that slow. So why isn't it continuously telling the pilot he IS stalled? (None of this about to stall warning nonsense.)

Methinks the damn good system software for the A330-200 can be improved, as can all software. And this is one good improvement. Enshrine the knowledge early so that when we DO get to autonomous SLF carriers this is one fewer problem the aircraft will face. (The communications problems should be covered as well with a "guaranteed" work around.)

galaxy flyer

All indications of AIRSPEED are based on air passing into the pitot tubes, processed by ADIRUs for output to the PFD or ISIS. Inertial is not used, and not usable, for airspeed.

One is not "dispatched" in alternate or abnormal law.

NWR

BEA knows exactly what the PF and PNF thought happened.

'At around ' 2h11min40 (why say 'around'?) the 'Captain re-entered the cockpit'

Hard to imagine him saying other than "What happened?"
And the PF and PNF will tell him and us what they perceived and their actions.

Mind you
The background stall noise on the CVR will be sounding like 0m55sec into this Tarom ROT381 AB incident:
perhaps it said 'At around ' because only then did he speak

JD-EE

matthewsjl, let's add one more "fact" to your discussion, The 60kt drop in speed is very close to what calculations show the plane would lose simply trading horizontal velocity for altitude with very high efficiency.

That's a good argument for there not being an updraft. And 215kt at FL375 is almost certainly qualifying as a legitimate stall, especially considering the angle of attack registered. (That's if I've read graphs right.)

jcjeant

Hi,

As many commented here ... the CVR transcript is not the complete one.
On the other hand ... Air France (in a press release statement) tell they were proud of the professionalism of the crew .. it's certainly for some reasons.
Sure Air France had access to the complete CVR transcript or sound version.
So I'm in the waiting of the words of the captain (no one word of the captain after the upset in the actual BEA release)

CONF iture

To me, 29% makes more sense than the 37% as initially estimated by the BEA in their interim report #1
Due to the heavy cargo load, it is probable that the fwd cargo hold was also fully loaded, which explain the fwd CG of 23% at take off. Considering the initial fuel distribution, the trim tank had normally 2400kg at take off time.
During the climb, an additional 2400kg was transfered to the tail for a total of 4800kg which was still there at the time of the event.

The 37% target is an ideal that can be reached only if the initial loading would allow it.

The CG was pretty much centered, at least it was not especially aft.

CONF iture

At the present time, who has, or momentarily had, access to the full or partial data ?

• some people at the BEA
• then who ... ?

grizzled

CONF...

Here is a link to ICAO Annex 13. If you look at Chapter 5, specifically sections 5-3 to 5-5, you'll find the criteria for who can be, who should be, and who must be part of the investigation, including who can have access to all or part of the investigation material.

http://www.airsafety.com.au/trinvbil/C619icao.pdf

jcjeant

Hi,

It's just for the organization between states ..... it's not realy instructive about who have really access to the data .... IMHO

mm43

JD-EE
Undoubtedly! It would appear that no-one anticipated an approach to a stall that wouldn't under normal circumstances be dealt with correctly. In this case it wasn't, and degradation of air-data was such that the SW operated in a negative sense, i.e. letting them think they were about to stall again, whereas if they had persisted they would have gone through the warning with a chance of recovery.

As takata has rightly pointed out, IAS greater than 60KTS AND/OR AoA less than 30 degrees are required to prevent the Abnormal Law activating. It is while in this AND/OR regime that a very clear warning needs to be given that the aircraft is stalled. IAS is useful, and AoA is useful, and it would be preferable to know each rather than being left in the dark because one or the other, or both wasn't deemed valid.

KATLPAX

"This is a stall. Reduce power and nose down!"

If this quote from the Der Speigel article is true, it seems a puzzling omission from the BEA report given the other quotes and seems to purposefully lead one astray(by omission). It was a definitive acknowledgement by the captain of their situation and how to react.

Graybeard

Wiki has some info on pitot-static, but does not explain. I don't think I've seen this explained elsewhere in a long time, so I'll take a stab at making it simple.

The pitot tube captures impact air, which is sensed by an Air Data Module in the A330, and converted to airspeed. The drain tube allows water to pass without clogging the pitot. To do that, it passes air all the time.

The pitot opening is a specific size, say 5mm dia, and the drain is maybe 2mm. That makes the pitot about 20 sq. mm area, and the drain about 3 sq. mm area, for a ratio of about 7:1. This means the impact pressure at the sensor is about 14% less than actual impact pressure.

When it comes to clogging at high altitude, there are these possibilities:

Pitot clogged, drain clogged at same time: IAS, Indicated AirSpeed, will not change with change of aircraft speed, but IAS will increase with increase of altitude.

Pitot Open, Drain CLogged: IAS will increase 14%. 14% excess IAS will remain regardless of actual airspeed or altitude.

Pitot Clogged, Drain Open: IAS will decrease toward zero as the drain bleeds off the impact pressure to ambient. Increased altitude will not affect IAS.

This last situation: "On 28 October 2009, an Airbus A330-202 (A330) aircraft, registered VH-EBA (EBA), was being operated as Jetstar flight 12... (IAS decreased to about 60 Kn). The airspeed disagreement was due to a temporary obstruction of the captain's and standby pitot probes, probably due to ice crystals. A similar event occurred on the same aircraft on 15 March 2009."

The AeroCaraibe incident showed much the same behavior, so it can be expected AF447 was also the same type failure.

Comment?

jcjeant

Hi,

Seem's a little bit ironical ... but modern military jet fighters planes have an HUD with many interesting datas just in face of pilot eyes .. all this for assure the safety of one soul .. and some tons of metal and weapons .. and airliners have not this for assure the safety of hundred of souls and some tons of metal.

fizz57

Just a general thought - does the Flight Path Vector display use air data at all or is it just inertial? Would this have helped the pilots' SA in this situation?