Of course many aircraft the same night did deviate, but given the likely weather pattern, to do so after 02:00 would surely have been too late? I didn't realize there was any doubt that AF447 departed from cruise altitude without a major incident - and I thought it was accepted that a '200Te falling leaf' would still take 4 mins to get to sea level? How long did the B-720 in 1963 take to fall from FL 370 before recovery at 14,000 ft, or the 1985 747 from FL410 to 11,000ft?

No-one can explain why it would have occurred after 02:10 under effectively 'manual control' and before would presumably require a QF72 style flight computer 'glitch', perhaps the pitot tubes are innocent?

But surely the assumed proximity of the LKP to crash site suggests one major unrecoverable incident?

HN39;

Just stepping out...but I can help orient - quickly, in the photo of the wreckage, we are looking fro the rear of the right wing towards the leading edge, (which is essentially gone), as you had discerned after my initial post, with the flap track to the right. To orient the BEA photograph of the track for comparison purposes, I "flipped" the photograph, then rotated it and turned it into a black-and-white. If one is capable of taking the original photo and doing the same in one's mind's eye, so much the better, but that would lead to some confusion when looking at these photos without knowing the whole process...sigh...complicated typing it, a moment to show! Sorry for any confusion. I'll help out further when back...

PJ2

Landing gear photo
 

This may help with orientation for the BEA pic "landing-gear-a":

Airbus A330 Landing gear | Flickr - Photo Sharing!

- that's one firm vote for a clean a/c.

My apologies to PJ2 and everyone and thanks to BJ-ENG for his excellent picture. My orientation was wrong, in the BEA report pics we are looking towards the wing leading edge on the right, the flap is left, and what I mistook for engine pylon is the wing leading edge. I have no problem anymore with the flap track in the wreckage picture.

Regards,
HN39

Quoth QFR "(hopefully the moderators will merge the two)".

Screams JD-EE NNNNOOOOOOOOOOOOOOOOOOOooooooooooooooooooooo!

(Phew that's a bunch of cranks, tools, and two or three jewels that should be quietly redirected over here. Let the people who think 911 was an inside job and that Obama is a Muslim born in Kenya stay over there.)

{O.O} (Yikes they're crazy over there!)

Shadoko asks some questions

"- if an engine flameout happened, was an ACARS to be transmitted?"

I can't say. But, I'd presume that would be fodder for a message. But past discussions make me think it might not be for some reason.

"- is a spinning "posture" making a stop to the satellite transmission (so, no more ACARS)?"

A spinning condition would not necessarily stop ACARS. They were in a position for a very high angle view of the satellite. The ACARS antenna has a very big "main lobe". So a mere spin should not take the satellite out of the main lobe of the antenna. It would, perhaps, show up in satellite logs as degraded signal to noise ratios (technically speaking Eb/N0 - energy per bit divided by noise energy.)

"- in the event of a vertical stabilizer rupture, is the cabin depressurization mandatory (and followed by an ACARS)?"

I believe you mean "inevitable." In the earlier thread there were pictures of the tail structure. Given the damage visible on the stabilizer it can come off in that manner without compromising the cabin. The entire tail structure probably cannot come off (violently) without compromising the cabin integrity. There is no indication that either departed the rest of the plane before the plane hit the ocean surface.

I hope the middle answer helps. That's more or less my field (my technical life's first love, you might say). The other two answers are "educated guesses" at best.

DFDR
 

What will the DFDR give out if it recorded erroneous data from the ADIRU ?

It will show exactly the same erroneous data... which will then NOT correlate with the other data..
Don't worry about that too much, milsabords, it's about the first thing the people analysing the data will be looking for.....

JD-EE writes...........
"Quoth QFR "(hopefully the moderators will merge the two)". (PFR)

Screams JD-EE NNNNOOOOOOOOOOOOOOOOOOOooooooooooooooooooooo!

(Phew that's a bunch of cranks, tools, and two or three jewels that should be quietly redirected over here. Let the people who think 911 was an inside job and that Obama is a Muslim born in Kenya stay over there.)

{O.O} (Yikes they're crazy over there!) ":uhoh:

Point taken PFR:ok:

milsabords;

BEA's Interim Report No.2 identifies the Make and model of the DFDR and says about the parameters: "The decoding document, supplied with this airplane, has around 1,300 parameters." It does not necessarily record the output of all three ADIRU's. Each ADIRU outputs airdata (airspeed, altitude, AoA, TAT and barometric vertical speed) and inertial data. Only airspeed and possibly TAT are likely to be affected by ice particles.

Gear down . . .
 

If I may, a brief musing on the subject of the gear being down . . .
The proposal we have so far from the BEA is of a very significant vertical deceleration at the time of impact with the water - as seen by the vertical crushing of internal spaces retrieved.
If you look at this another way, the gear does not know at that instance that it is meant to be decelerating as its mounting effectively means the major part of its mass is still in free-fall - so why would it not continue downwards? As has been said, the locks are not designed to resist this sort of force.
Thus it is likely to release at the point of impact and continue to deploy under gravity, avoiding the torsion of a water impact at speed.

Any sense?

HN39;
...not that it's important now but what appears to be the leading edge on the left side of that photo is actually the next flap-track canoe, but your orientation is still correct. I had the orientation wrong in my initial post in answering your original question, thinking we were looking rearwards, at the left wing/gear with the entire leading edge gone. What turned out to be the flap track, I thought was what remained of the left pylon after the sheet-metalwork had been stripped off the support structure. Anyway...it just shows how careful one must be in drawing conclusions even when one knows the airplane.

On the recorders, parameters and what's recorded, the other aspect of this is, though it isn't mentioned in the BEA Reports, this aircraft will also have had a Quick Access Recorder of some sort, for AF's FOQA Program work. The QAR isn't crash-protected and is usually in the EE Bay below the cockpit (for this type). It's source of data are the same (ARINC) buses that feed the DFDR, usually through a DFDAU, (digital flight data aquisition unit) or FDIMU, (flight data interface management unit).

Whether the card or other recording medium and its electronic housing survived the impact and the time underwater is of course an open question, but it IS a source of data which in all likelihood will have many more parameters, (upwards of 2000 - 3000 parameters), and at different sampling rates, than the DFDR.

Time will tell, but I hope there is included in the recovery plans, a search to determine if the QAR recording medium can be found.

Alison;
Completely.

Debris Location and Original Storm System
 

Has anyone overlaid the new debris location with the elegant weather details that Tim Vasquez figured out 2 years ago?
It would be interesting to see how they relate to each other......

Unlikely scenarios...
 

Quote from takata (post #3093):
a) How long did they maintain cruise altitude?
It is still unknow without the recorders. It seems that they may have lost a lot of altitude at the end of the sequence (02.14) when the pressurization valve opened. An emergency descent up to 8,000 ft/mn is plausible or it may be due to a previous but recovered upset.
b) a gently stalled aircraft from cruise level would take about 10-15 minutes to reach sea level, not 5 minutes, and it would certainly cover a fair distance from LKP if not circling around.
c) At impact time, this aircraft seems to be in controled fligth attitude... if powered. This may be due to a failed attempt to relight its engines without having enough altitude in order to come out of the dive to build up speed: a CFIT at the dive bottom rather than an upset from cruise lasting up to this point.
d) an attempt to ditch seems very unlikely.

Salut, I think you were giving consideration to less-likely scenarios, even though you agree with me that flight beyond about 0215z is unlikely?

So I will try to join your discussion, further to my earlier comments:
"If they had maintained cruise altitude during that 4 -5 minute series of failures, and then had a simultaneous double-engine (and therefore double-generator) failure, they could indeed have glided for about 12 -15 minutes. Seems a bit odd that they should end up ditching without pre-warning cabin crew and passengers, and find themselves within about 15nm of where they were when their problems began."

Just to remind others, the absence of any received ACARS messages after 02:14:26 (at least one was due) is most likely explained by complete loss of AC generator power. Now: back to your above-quoted points.

In (a), you recognize that the most likely explanation for the Cabin VS warning at 0214z is that the aircraft was descending rapidly through about 6000ft, with the inward relief valve opening because the cabin altitude was >6000ft. However, a simultaneous double-engine failure at cruise altitude could produce the same warning until the outflow valve had a chance to close to maintain cabin pressure.

In (b), a "gently stalled aircraft" for over 10 minutes from cruise to sea-level seems a big stretch of the imagination. As you say, it would also involve a circle or tear-drop to get back close to the LKP.

In (c): with the high probability of a double-engine failure leading to a ditching, the flight crew would know that the cabin crew must prepare the passengers accordingly. The fact that this was not done suggests there was insufficient time, and/or the cabin crew were unable to reach either their interphones or the cockpit, due to an upset.

In (d): if the PF had regained control, but neither engine was running, he would be doing his utmost to achieve a controlled ditching. But we know that the vertical speed was still enormous. The PF may or may not have been recovering from an even higher VS.

I still go for loss of control between 0210z and 0211z, possibly as they exited the Cb.

KTVaughan, Didn't Tim Vasquez mark the LKP on his graphics?

Chris (with apologies to Bobman84...)

Debris Location and Original Storm System
 

Not the info from Tim Vasquez, but I have added an image from BEA's track plot for 02:14:30 to my Googleearth layers for AF447. The dark orange part of the storm nearly fills the 40nm radius search circle. The BEA track/weather images are here:

Flight Paths of Flight AF 447 and of the flights that crossed the zone around the same time

Note the scale of these images: the coast of Brazil is in the lower left corner. The dashed lines through the center are the equator (0.0 lat) and 30 deg W longitude. I assume the weather is from a satellite image; I don't know what the colors represent (maybe cloud temperature).

I have cut a section with an overlay of the 40nm circle (in the grey box):

http://i51.tinypic.com/2zofar5.jpg

There are some indicators that will help, e.g.

1. Those bodies recovered were probably not strapped in at the onset of the upset.
2. Some abnormal 'g' forces were experienced throughout the upset period.
3. A major pitch up in heavy turbulence could have resulted in a rapid negative CG shift - unbelted cabin crew and pax thrown toward the rear of the cabin.
4. Any ensuing spin/rotation would have introduced forces that prevented cabin crew and pax returning to their seats.

In short, the longitudinal stability at some point became compromised.

Concur fully. That is why I've been looking at unpopular concepts like dynamic loss of control recently. It even fits the rudder limiter position as discovered which most scenarios do not explain.
Dynamic loss of control is rooted in servo theory and relates to PIO.
It appears that if one does not work with stability and control on a routine basis, it is as easy to comprehend as Ancient Greek.:rolleyes: Come to think of it, the formulas are full of Greek letters.:}

The flight control system should be reconfigured to ALT, but for another reason than a double ADR fault. A double IR fault would maintain the possibility for "low speed stability" to activate.

Call them any name you like but not before you study building 7 ...

Do they mention the slats ?
It would be interesting if you can see anything in the pictures about the slats as it is part of the procedure to deploy them in an attempt to exit a stall situation.

Machinbird, it's a little frightening as animated meat cargo to think that a plane has discovered a new set of conditions that leads to oscillations within the feedback loops in the control systems. (That is what I read your "dynamic instability" to mean.)

I suppose it's possible if somehow the ambient conditions around the plane were very much different than expected or the set of feedback parameters that existed to work with were so corrupted nasty results happened. Of course, for much of this crash what is often called "meatware" was in control. And it's properties are not precisely calibrated, especially if conditions were quite different from any training the "meatware" ever had.