Vertical Stabilizer - in drift mode!
 

I am mindful of a video showing the recovery of the V/S. The rudder is clearly jammed at around 30 degrees to starboard, which is consistent with the V/S initially being propelled off its clevis joints in a forward and to port motion. It will have "belly-flopped" into the water on its port side and the rudder will possibly have been rammed over to starboard at that point.

The torn off fuselage skin on the starboard forward side will also have had a part to play and I believe the V/S was moving through the water along a line drawn through the fuselage skin piece to the top and aft end of the rudder, i.e. with the rudder leading the way. On top of that the V/S including the rudder is a large and light aerofoil (8.5m [base] 9.3m [high]), which in this case will not take much wind to get it moving and it will "sail" along quite nicely. The effective waterline length was around 13m (LOA 15m), with the top of the V/S pitched up a few degrees, and the hydro-dynamic drag would not have been high as the base was open to ribs #2 & #3 and didn't present a solid bulkhead.

We don't need the "fat lady to sing" - and hopefully never!http://images.ibsrv.net/ibsrv/res/sr...ies/thumbs.gif

mm43

Revisiting Simple Solutions
 

Three weeks after the accident, Takata plotted the reported position of recovered items and concluded: "...even if the drift was actually half my figure (it may also be twice), AFF 447 would have turned back and not crashed close to 2014Z." In December, 2009, mm43 charted the expected influences of the North Brazil Current, the North Atlantic Equatorial Current, and the North Atlantic Equatorial Counter Current and noted the conclusions of Météo-France, SHOM, IFREMER, Mercator Ocean, CROSS Gris-Nez, the U.S. Navy, Brazilian Navy and U.S. Coast Guard (USCG) as well as his own. A lot of talent has been focused on determining where to search for the hull. A lot of seabed has already been covered testing many hypotheses. So far, no joy.

By way of revisiting the thinking:

2 July, 2009 BEA Findings

If the A/C impacted the water intact in a straight line with high vertical acceleration (stated as a fact by BEA), then breakup of the fuselage and ejection of debris occured at that moment, meaning that on the morning of June 1st, debris and bodies were colocated in a small area, likely less than a kilometer in diameter. Destruction was so complete that wreckage and bodies from all parts of the A/C interior, fore to aft, were later found. For the following six days, until first discovery of wreckage and subsequent recovery, each item drifted in a way determined by its windage, buoyancy and the environment. According to BEA's 1st Interim Report, a database of 640 discovered items was built, which included the date/time recovered and lat/long. Later, in the 2nd Interim report, that database was described as containing about 1000 wreckage items (not including some 50 identified bodies).

Appendix 4 of the 1st Interim Report is a graphical interpretation of the 640 found item database. While these BEA graphics lack granularity or information detail, it is visually apparent that the majority of bodies were discovered in fairly well-grouped clusters on June 7, 8, 9, and 10. It makes sense that the bodies would be clustered: unlike various aircraft parts and high-floating debris, they share similar properties of form and would respond similarly to local current, wind, and sea conditions.

Likely, the dispersion and separation within this once tightly-grouped cluster increased to perhaps several square miles or more over six or seven days. So search and recovery needed to cover a large area each day, even when the general location of the cluster was identified. Each day, according the the BEA graphics in Appendix A, the cluster drifted almost due northward in a straight line along the 30º30'W meridian, about twenty nautical miles per day:

June 7: 3º 45'
June 8: 4º 05'
June 9: 4º 28'
June 10: 4º 45'

http://i958.photobucket.com/albums/a.../bodyDrift.jpg

The Google Earth KMZ file has been updated here.

So for about four days, June 7 through June 10, we DO know what the current was doing at a pretty steady pace: about 20nm per day due northward (with a tiny bit of westing). So heuristics (a rule of thumb nicked slightly by Occam's Razor) might cause us to ask "What would be significantly different during the preceeding six days in that patch of ocean," and point us to an ultra simple linear solution for the initial six days of drift: about 20nm per day x six days = 120 nautical miles, backtracked almost due southward from the June 7 body cluster.

This solution is waaay outside of BEA's "40nm from last known position" search radius and its "lawn mowing" and occasional Phase Three diversions. Understandably, there are tricky currents in those equitorial waters, as the BEA Report clarifies:

But nobody was there to actually measure the set and drift in that part of the ocean and verify the computer models, checking the probable variations within that hundred or so mile radius patch of interest. We do know for a fact, however, that on days six, seven, eight, and nine the cluster of bodies was setting almost exactly northward, drifting at somewhat less than a knot, with windage and wind direction playing a minor, if not negligible, role (particularly in comparison to the VS and the Crew Rest compartment). Perhaps I am missing something here, like charted and known counter currents and swirls, surface storms, or other variables in this area occurring between June 1 and June 6... In any case, sometimes computer models can become overly complicated.

There are so many unknowns!

As Phase Three of the search effort concludes, some very hard and possibly expensive decisions will be made.

GB

I have seen reference that the Brazilian Navy dropped marker buoys very early in search, so some real surface drift should be known.

The expected currents were in this post, which partly explains the search area

mm43's post

But your suggestion is remarkably similar to that suggested a long time ago - do check the date posted

takata's post

I wonder if it is now possible to map where the wreckage is not - the original visual search will have oveflown the expected flight path etc...

Ghostbear,

interesting and reasonable analysis.
I'm wondering anyway why not more seems to have been published by BEA on drift patterns.
Moreover also the Search patterns seem to ignore them largely.
given the trend between 6th and 17th of June, why would anyone perform a detailed search north or even at the poistion of the debris / bodies of June 6th.
From my (very amateur) perspective starting around the location identified by mm43 and then going e.g. in concentric circles would have increased the likelihood of finding something significantly. At least much more likely then around the location of the bodies found on June 6th. And I guess most people around here would agree.
So many experts and big dollars being involved I'm a bit uncertain what us mortals are missing?

40 nm RADIUS FROM LKP − IS IT ENOUGH?
 

GreatBear suggests above that the crash might have been outside the 40nm radius from LKP (last-known position), on the basis of back-tracking of body-drift. In response to takata’s comparable “Probable Crash Zone”, posted here on 21st June 2009, FlexibleResponse stated here that “you also need to take into account that the human body does not float in seawater until about 72 hours after death. Then there should be some allowance for the time from establishing positive buoyancy to float from the wreckage at the bottom of the ocean up to the surface. Again some allowance should be made for the average sea current during the ascent time…” (etc.)

Not sure if FlexibleResponse’s assertion has been challenged since but, if not, it put the crash zone back into the circle of 40nm radius. So what about the limited evidence we have of possible flight-paths after LKP, based on evidence of ACARS messages and BEA analysis of aircraft debris?

At risk of re-hashing previous discussions, the implications of the BEA Tests and Research into ACARS messages (Para 1.16 of Interim Reports 1 & 2), as I interpret them, include the following:
02:10:10z (“AP-OFF” message received from ACARS) indicates aircraft at, or close to, planned cruise F350/M0.82, GS<480kt;
02:10:34z (routinely received from ACARS) LKP − FL and GS unconfirmed (?);
02:14:45z (inferred from absence of receipt of an expected ACARS transmission) − flight has ended;
aircraft was intact until impact at sea-level, and impact GS (not stated) unremarkable, so probably below 300kt.

So one may infer that the aeroplane descended 35,000ft in not more than 4min35sec, starting with a GS of, say, 480kt (8nm/min). In addition, it had to reduce GS to something below 5nm/min prior to impact.

This suggests a rate of descent averaging 7600ft/min. That is nearly three times the average VS obtained at normal descent speeds (prior to deceleration) with idle thrust, yet TAS seems to have been considerably lower at impact than at start of descent. Where did the drag come from; to dissipate so much potential energy, and some kinetic energy, so quickly? Spoilers/speed-brakes are unlikely to have been used, and there is no evidence that the L/G was ever extended.

[If the ADR used by the A/THR was grossly under-reading the IAS/IMN, thrust above idle may have been maintained − despite descent − until 02:10:47z (ACARS “A-THR OFF” message received), or even later if the throttles were left open at that point (thrust lock).]

There’s been speculation that the loss of FPV (02:12:10/16) would have been caused by TAS>599kt (10nm/min), although this datum would be from an ADR that had been delivering under-readings of IAS previously. If the relevant pitot had thawed out in 2min, for reasons unclear, excessive TAS remains a possible cause; although the “measured CAS lower than 60kt” may be more consistent with the likely scenario of a blocked pitot tube, coupled with rising static pressure (in the descent) from unfrozen static ports. The third possible cause of the FPV being disallowed, VS>20,000ft/min, is implausible on an intact aeroplane − unless previously frozen static ports were unblocking.

Overspeed - A/C Intact

Although it is possible that the aircraft stalled at high altitude, it seems unlikely with airspeed indications under-reading. Overspeed is more probable, particularly early-on. From the BEA’s comments on condition of the control surfaces recovered, this would not have been supersonic, so that limits average descent TAS to about 600kt (10nm/min). ** In this scenario, with a clean aeroplane, the speed reduction from over 600kt to 300kt − albeit achieved during a shallow descent − would have taken nearer two minutes than one, I suggest. That reduces the average TAS to a figure well below 10nm/min, and an average GS of a similar order. The alternative deceleration scenario of flare into a steep climb, followed by deep stall, would lower the average TAS further, presumably.
(Note: all this is predicated on absence of in-flight structural failure, and an impact time not later than 02:14:45z.)

Ignoring the unpredictable wind components, which are not likely to have amounted to much overall, it therefore seems unlikely that the aircraft would have travelled more than 40nm in the 4:11 after LKP, i.e., averaged a GS of more than 574kt (9.56nm/min). And, even if it could, is it likely to have travelled in one straight line, i.e., continuing north-eastwards along the airway?

Given the above, and inferring from the evidence that flight beyond 02:14:45z was unlikely, the 40nm radius of search from the LKP (last-known position) seems to suffice.

Chris

** If the descent stabilised at, say, M0.95, the IAS/CAS would rise rapidly through VMO; the TAS rising at a much lower rate. Considering the BEA has not reported tell-tale signs of extreme dynamic pressure to the tail fin (V/S) and flying controls retrieved so far, one must assume that the descent TAS at lower altitudes would have been considerably lower than 600kt. No doubt, various computer models will have been run by the BEA, not to mention trials on flight simulators.

Update: "Seabed Worker" - position
 

An apology for the positions last posted. It transpires there were transcription errors in the positions, and the "Seabed Worker" didn't go out to the west of the current search zone. That post has now been removed, and the correct positions over the last two days follow:-

23 May 2010 16:20 Hdg 276.5 Spd 00.6 3°05'42"N 30°45'08"W
23 May 2010 07:11 Hdg 255.9 Spd 00.8 3°04'59"N 30°41'40"W
23 May 2010 05:30 Hdg 101.1 Spd 10.3 3°06'41"N 30°49'52"W
23 May 2010 03:45 Hdg 303.9 Spd 01.0 3°10'42"N 30°49'13"W
22 May 2010 19:00 Hdg 082.3 Spd 07.2 3°28'58"N 30°34'25"W
22 May 2010 17:09 Hdg 024.5 Spd 10.3 3°21'09"N 30°40'13"W
22 May 2010 06:34 Hdg 261.7 Spd 00.4 3°07'10"N 30°41'35"W
21 May 2010 18:18 Hdg 323.5 Spd 00.8 3°07'16"N 30°41'11"W

http://i50.tinypic.com/207u79i.jpg

It would seem that the vessel has been and had a look at some areas to the north previously covered by the "Anne Candies" with her towed sidescan sonar. Currently, she is back in an area around 10NM northwest of the LKP.

mm43

MM43's apologies above should really be mine for having passed him dud info. So, apologies to all.

The search grids, June 1 through June 5
http://i136.photobucket.com/albums/q.../SNAG-0297.jpg

The search grid on June 6, when the first wreckage and bodies from the plane were spotted.

http://i136.photobucket.com/albums/q.../SNAG-0298.jpg

Unfortunately, in retrospect, the searches June1-5 concentrated on areas to the right of the track, and south of the last reported position.

In the June 1-5 image, the June 1 search grid was the narrow rectangular box from the last reported position almost to Tasil. The June 2nd search grids were mainly the two rectangular boxes, almost perpendicular to the June 1 search box, and to the right of the track.

I believe the circle in the June 6 image was not a search grid. but only an indicator of the distance from the last reported position and the debris sightings.

Where to now?
 

At this juncture it is important to look carefully at the proposals put forward by Great Bear, takata and also observations made by Chris Scott, sensor_validation and others, because each submission has a point or two that could add to the understanding of "where?".

From my own perspective, I have possibly become blinkered through making use of the OSCAR and QuikSCAT satellite data and trying to make it fit a solution. sensor_validation pointed out an early post of mine which attempted to make sense of the ocean current circulation at the time of the crash. What it does indicate, is that the aircraft crashed in a fickle part of the equatorial North Atlantic, and that available current data can not be relied upon explicitly. Some other factors, notably drifter buoy data from the surrounding area, once again tend to confuse the issues.

Getting back to the initial airborne searches conducted by the Brazilian Air Force (FAB); they also need to be carefully considered, though on a couple of days the cloud base was relatively low and visibility marginal. It could be that the debris was overflown, but one wonders on the lack of any radar targets. Composites don't make good targets, and neither do spoilers etc. floating flush with the surface. So the lower the altitude flown during a radar search, the less chance you have of getting meaningful returns - other than sea clutter. The following is a screen shot taken in the Rescue Coordination Center at Recife showing areas that have been searched by the FAB, and from it you will note the limited area to the west of the LKP that was covered.

http://i47.tinypic.com/5nljeu.jpg

Later screen shots [not shown here] show an extensive grid search undertaken once the initial debris was sighted.

In short, where we are now provides a very good and expensive analysis of where the aircraft didn't crash, and a rethink and further examination of the available data is a "must" before further funds are committed to another search.

mm43

Based on mm43's positions in post: http://www.pprune.org/5710975-post1125.html, I am puzzled by the northern excursion on 5/22. Based on the times of the preceding and trailing positions at the south westerly search area, they spent less than 20hours at the northern site, possibly much less. One of those positions is on the eastern edge of search zone 2 of BEA's 5/4 extended search areas. <20hrs is not sufficient time to deploy transponders, launch and recover the AUVs and recover the transponders, and also have any useful bottom time. Zone 2 is 16 x 3.5km and would take about 16-18 hours to search using both AUVs, based on the search rate I proposed in post: http://www.pprune.org/5692187-post1029.html. Note that at both these positions, Seabed Worker was travelling at a transit speed, and was not loitering. I wonder what that excursion was for.

I find it interesting that on 5/23 Seabed worker appears to have started a new search area, as noted by mm43. I thought they were due in port at either Recife or the Cape Verde Is. on 5/25, and would have had to leave the area no later than the evening of 5/22 to make either transit, assuming a 12 knot cruise speed (http://www.seabedgroup.no/attachment...WORKER-007.pdf).

Edit: I re-read the BEA announcement of 5/4, and it says "The search operations in the zone will be completed around 25 May." So I guess the 5/25 date is for the end of the search, not the arrival in port. Sorry for the noise.

Based on that observation, I'd say you are right. As to the deviations to the north, and what was achieved, I am in the same position as you and wondering.

mm43

AF447 dead treated like a footnote by Air France
 

Ben Sandilands, Australian aviation journalist and blogger reports: AF447 dead treated like a footnote by Air France.

HN39, waiting the FCOM value of V-alpha-max, Vs1G~M 0.56 (at FL350, with 0° roll, M~205t and clean config) and as a rule of thumb, Vstall~0.93 Vs1G~M 0.52. But you are interested by the value of Vstall at sea level ? (secondary stall/overcorrection at low altitude in the absence of alpha prot ?).
Back to the search area,
-the initial visual searches (SaturnV reminder): if no debris or body was found during the 5 days after the accident, it is probable that the search means were not deployed on the right spot: if most of the initial searches were dedicated to the south and the east of the LKP without any results, it suggests that the debris/bodies were initially located to the north or west of the LKP, away from the AF 447 expected route.
From the debris/bodies time distribution and back drifting analysis: there are two groups of results, one from the Meteo France (Mercator model) which suggests that the debris/bodies were drifting with the north equatorial currents (oriented to the NE at 2° north and to the NW at 4° north) and a group of results from the US Navy, the USCG and the Brazilian Navy (using the same NCOM currents model) suggesting that the debris were initially under the influence of the north equatorial counter currents or combined influences of the two currents (north at 3°N, rotating to the NE at 4°N). The latest recovered debris drift clearly showed a NW rotation of the currents, consistent with north equatorial currents: it would suggest a back drifted estimate closer to the Meteo France or MM43 estimates, largely (between 25NM and 40 NM) to the west of the LKP (see the MM43 analysis from NOAA/OSCAR current maps).
-the one month pinger searches: suggest that the underwater beacon locators (assuming at least one of the UBL was still functionning) are not to be found to the north of LKP (explored using TPL with a good confidence level). But the French Navy helped by Thales (pingers waveform detection) suggests that the UBL could be located in the south-west area of the 40 NM radius circle, little bit (10-20 NM) south of MF/MM43 estimates.
-from the areas already searched using the sidescan sonars during the phase 3 without success (mainly NW of the LKP) and the above considerations, the W-SW areas of the 40 NM radius circle seems by far the best candidate now for further searches and I really hope that a new search phase will be funded for this SW area.
Jeff

Chris Scott: Thank you for your analysis. Descent of the aircraft from normal cruise at FL350 to sea level in less than four and a half minutes presents a nightmare of darkness, bad weather, unusual attitudes and perhaps overwhelming g forces. As you point out, the timing between 02:10:47z and the unreceived 02:14:45z ACARS messages "suggests a rate of descent averaging 7600ft/min. That is nearly three times the average VS obtained at normal descent speeds (prior to deceleration) with idle thrust..."

So many unknowns. Might the aircraft have been still in the air beyond 02:14:45z (asymetrical flame-outs -- loss of power)? Uncontrolled, would this 205t aircraft "drop like a stone?" Had the crew already turned back (they were west of track at 02:10)? Many questions already discussed in this and the prior thread. Few hard answers.

But all might agree now with Hyperveloce that "the SW of the 40 NM radius circle seems by far the best candidate now for further searches and I really hope that a new search phase will be funded for this SW area."

While there have been beneficial lessons already learned from this tragedy, it seems there will be no definition without more data. Mow the lawn! Readable black boxes are the holy grail.

As we approach the first anniversary, my thoughts are with the many relatives and loved ones who have lost so much.

GB

Had it not been determined that 447 was tracking left? Also at least a few had proposed that she had made a (complete?) turn? Had the uA/S wacked the panel sufficient to dispose a turn back? If that is not the case (apparently) the virginal ocean bottom W(NW) of LKP doesn't deserve a look, but if otherwise?

mm43, Hazelnuts, Hyperveloce, etc., you are to be commended for your persistence and scholarly work. There is no corporate agenda that holds a small candle to the need for blazing light on this operation, for the good of all.

Chris,
what 'tell-tale signs' do you have in mind?

HN39

Please excuse the interruption.Will Fraser and others immediately brought up LOC and loss of airframe integrity. The hasty First Report was worded such that these possibilities were impossible; "En Ligne de Vol", "Intact at impact" etc. BEA has assumed from the outset an adversarial and defensive posture, imo, one that attempts to undermine even a glance at another "conclusion" than "Not enough A/S for level flight, but 'oh so close'".

The aircraft is defended in a sly way, since the "conclusions" are premature, openly invite disagreement, (with evidence), and properly belong in the Final.

rgards, bear

"intact at impact" you could say that of the recent puzzel at tripoli, definately intact, but a nanosecond later and most of it you put in your pockets
perhaps thats why they cant find anything, the bits are just to small.


gs

Yes, GS, but some bodies were found. If the plane really broke into tiny fragments on impact, what are the chances that whole bodies would have survived (and the VS also) ?

Ah yes, the bodies. On the one hand, there are 43 which exhibit injury consistent with seated and vertical acceleration. But wait, werent there 53 bodies recovered? What of the others? Uninjured? Or was their injury inventory inconsistent with the determination of BEA's that there was no loss of Hull integrity at altitude? No matter, even though the list of 43 victims offered "support" to BEA's "conclusion" it is on the surface only. A human ejected into the slipstream of an airliner's at cruise would exhibit these specifics almost certainly. So there is no academic value to BEA's assertion. What then? If it proves nothing, why include these, exclude the other results, and 'conclude' anything at all? Unfortunately, most of the readers who have ingested the BEA reports, assume BEA is correct. They could be. But it is in no way certain, just as any conclusion would be, gelled prior to conclusive evidence.

BEA, Air France, and ABI all have a monstrous stake in the findings, here, and the findings needn't be accurate, they need only be believed. So, what do we have? A sales job. Not even a good one. No matter. In their shoes, I would be claiming all the same "conclusions".

HN39,

Me: “Considering the BEA has not reported tell-tale signs of extreme dynamic pressure to the tail fin (V/S) and flying controls retrieved so far, one must assume that the descent TAS at lower altitudes would have been considerably lower than 600kt.”
HazelNuts39: “Chris, what 'tell-tale signs' do you have in mind?”

Neither a structures man, nor an accident investigator, I am admittedly making at least two layman’s assumptions: that there would be tell-tale signs; and that the BEA would have no reason not to report them. So it’s what they have NOT said that leads me to infer that the aeroplane was not flown at, say, more than 100kts over VMO. Like yourself, I haven’t got access to the FCOM at present, but no doubt the VMO is below 400kt (about 350?). For the purpose of crude argument, such as mine, precise figures are pointless.

Quotes from Interim Report No.2:
“[caption to photo] Upper surface of left-hand inboard aileron with the fittings attaching it to the wing aft spar: failure due to the bottom-upward loads applied on the aileron [at impact].”
“On the right-hand elevator, four of the seven [attachment] fittings were present. They had bottom-upwards deformations.”
“The vertical stabilizer was in generally good condition. The damage to the side panels was largely due to the recovery… These observations indicate that [the V/S] was subjected to a load greater than 120,000N in the rudder’s hinge axis.”

So the condition of the various flight-control surfaces recovered is apparently not worthy of comment. The tail-fin damage is impact-only, bottom-upwards.

Bonne fête!

Bearfoil,

Most of the arguments here are based on the BEA report, and some on their provisional “Findings” (there are no “Conclusions” in their Interim Reports). That does not necessarily mean that we accept all of them unreservedly, or that the BEA will not amend − if and where necessary − in the light of further evidence. But we need some of you to play devil’s advocate!

Chris

- no, as far as I am aware.

As to BEA report LRP has been in fact slightly left of track.

Flight beyond 02:14:45z
 

GreatBear,

The problem with flight beyond 02:14:45z is that − as was pointed out to me recently − unless the ACARS transmission system coincidentally became u/s (aerial damage, or plain bad luck), or its aerial lost sight of the satellite (unusual A/C attitude), the first engine (and/or its AC generator) to fail would itself have resulted in at least one ACARS message. By the same logic, there needed to be an almost simultaneous failure of both engines’ AC generators to disable ACARS transmission of the first such failure.

If the BEA is right, the absence of the predictable FCPC1`(PRIM 1) FAULT message at 02:14:45 can only be explained by one of the above events having taken place between 02:14:26 (Advisory, Cabin VS) and 02:14;45.

Like you and others, I share Hyperveloce’s concern that the SW quadrant of the 40nm circle should be searched, asap. But we're not paying...

Chris

Well Chris, FlexibleResponse's assumption of 72 hrs after death for body surfacing is not valid at sea into such context:
A very large majority of people are naturally buoyant into sea water while a majority of them will also sink into non salted water. Moreover, people which are not naturally buoyant will sink into such a great depth that they would possibly never surface again due to very high pressure (slow decomposition -very low temp- would not overcome such pressure). Remember that the ocean floor is 2,000-4,000 meters deep around this zone.
As a matter of fact, all models used for locating the wreckage (US, Russian, French, etc.) were considering that the recovered bodies drifted at the surface from the crash time.
S~
Olivier

Bearfoil: Not sure if you got your total body count from my typo. I had typed 53 in an earlier post when it should have been "50 identified bodies" (per BEA 2nd Interim Report). Post corrected.

Chris Scott: Thank you again for your clarity. You would improve any logical analysis. Wish we had a wall chart showing weighted probabilites and those little triangular If/Then icons. Would need a large wall. Would be dense with facts and maybes.

GB

Hi,

Le Figaro - France : Les recherches de l'AF 447 virent au conflit d'experts

The research of AF 447 saw the conflict of experts

INFO LE FIGARO - The third season of research should be completed Tuesday morning on a failure. The BEA and the army opposed to the position of the wreckage.

The drama of the AF 447 is a mystery for now. The third season of research should be completed Tuesday morning on a failure because the investigators are currently on route Recife in Brazil without being able to locate the wreckage. Yet, despite two months of operation in the South Atlantic, the uncertainty has never been greater, as the experts disagree on the position of the device.

Background. On April 28, investigators returned to Recife to BEA after one month of research at sea after a crew change, they go back to finish searching the area for the third campaign. This area of 2500 km2 was defined from the study of drift of aircraft debris and the bodies of victims recovered in June 2009. A week later, once the theater. The army announced that it detected the signal from the black box: it is a distributed signals heard in June by the nuclear submarine, the Emerald and managed to positively identify the signal emitted by the black boxes of ten months AF447 earlier. "We sent twenty Emerald positions when the signal was seen black boxes," says one close to the matter. This new area south of the last known position of the device indicates that the Air France flight was turned around before crashing into the ocean.

The Seabed Worker, who participates in research 40 nautical miles further north, was immediately dispatched. It passes the new area in detail. In vain. "There was nothing to find in the south, said a member of the investigation. Everything has been raked. "Seabed The Worker then finish off again to explore the area of the BEA without broadening the scope defined by the army. "They spent 24 hours and no more on our area," says a source at the Ministry of Defence. The ocean is vast and we are the only ones to offer a scientific data and tangible. "

Reportedly, the Army has conducted tests in the Mediterranean with black boxes to identify their signal. According to a source close to the BEA, focusing on sounds recorded in June 2009, the Army had confused signals trials in the Mediterranean and those of the AF 447. She would have indicated an area that was not good investigators. "There has never been south of 447 AF, we lost one week, said a relative of BEA. The Navy wanted further work before announcing anything but the Ministry of Defence wanted to communicate quickly and we cut the grass under foot. "

"Explanations unclear»

Since then, relations between the BEA, under the Ministry of Transportation and the Ministry of Defence will be a bit cold. "The explanation of the army are unclear insists near BEA, and since it's radio silence on the subject. Asked by the Figaro, the army has not this story:" We are The third deterrent to the world and we are not here to please public opinion, says a source at the Ministry of Defence. We have removed some positions, but we maintain that certain positions are communicated to BEA correct. They have not been explored: BEA prefer to say that the army was wrong rather than saying that we received the signal from the black boxes but no one may know where he ever comes. "

The propagation of sound in water environment is indeed very complex. The black boxes of the Ethiopian Airlines jet that crashed last January off the coast of Lebanon were heard at 16 nautical miles from the crash site while recording only emit 5000. "If nuclear deterrence is hiding under water, is that this environment is complex," said one source at the Ministry of Defence. Today we have the probabilities of position of the wreckage. '

A fourth round is planned

The failure of this third season does not mean that the wreckage of flight AF 447 will never be found. Indeed, a fourth phase of research could be organized after-school new data from the army and the drift models of aircraft debris. The BEA has been able to study up on the drift and returns with data finer than the models that have so far been exploited.

"It is sufficient that the BEA was mistaken from 0.2 to 0.3 knots (about 0.4 to 0.6 km / h, Ed) for our area is the same," says a source at the Ministry of Defence. The echoes of the AF 447 will also be re-analyzed. The definition is complicated by their position on terrain and the phenomena of echoes. Asked by Le Figaro, a government source said "there is little chance that this new campaign, if it occurs, should be held before next autumn."

Great Bear,

As Chris Scott has said (eloquently so), precision is unnecessary thus far. Not a conscious mistake, but I was using your number. My purpose in noting that there were more identified remains than the 43 mentioned in BEAII was to illuminate a device BEA have used, and committed disrespect for those unmentioned. Only those injuries considered useful to support the conclusion of 'seated, vertically accelerated' were noted, and those of the others not mentioned at all. This is a callous act in my view, again, too soon for conclusions, and to present evidence at all, before all hope is lost of recovery of the recorders is precipitous.

More examples of minor propaganda? Where is the photograph of the spoiler? Was it left out due to its marked difference in the sequence of destructive forces at work?
Where is the opinion on the Leading edge collapse of the V/S? With sufficient Lateral energy to fail the V/S on AA587, it too looked in rather good shape after it fell into the sea from altitude. Compare? Contrast?

Why is the Vertical take up arm featured as substantive re: Vertical failur? Look at the environment in the photo closely, The end of the arm failed in the vertical, but the Hinge is undamaged? The "Lateral Rods" between the V/S paired mounts show failure, but look closely, the failure was not in the vertical at all, but in the designed for direction, laterally. Why is the attachment point for the V/S a tab of Resin inside a saddle of Steel? Two Phase does not do focal Stress well at all, but makes up for it in its resilience when Stress is "Spread". Why are the Lateral rods attached to the same rail? With a stagger, they could be twice their length, and save the V/S in case of critical lateral loading, even past the failure and loss of the Rudder. Ah, Weight and Money.

At some point I'll attempt to bring together the thoughts of the devil's advocate.
Every table set for investigative thinking and discussion should have at least one maverick, no?

all the best

bear

Update: "Seabed Worker" - position
 

The latest available positions show the vessel working in an area previously covered by "Anne Candies".

24 May 2010 16:35 Hdg 134.2 Spd 09.4 3°30'07"N 30°35'28"W
24 May 2010 05:57 Hdg 178.4 Spd 04.0 3°32'54"N 30°32'32"W
24 May 2010 04:16 Hdg 290.9 Spd 01.3 3°34'02"N 30°35'40"W
24 May 2010 02:59 Hdg 294.4 Spd 01.1 3°33'25"N 30°34'16"W
23 May 2010 17:46 Hdg 021.1 Spd 10.6 3°19'31"N 30°40'46"W
23 May 2010 16:20 Hdg 276.5 Spd 00.6 3°05'42"N 30°45'08"W
23 May 2010 07:11 Hdg 255.9 Spd 00.8 3°04'59"N 30°41'40"W
23 May 2010 05:30 Hdg 101.1 Spd 10.3 3°06'41"N 30°49'52"W
23 May 2010 03:45 Hdg 303.9 Spd 01.0 3°10'42"N 30°49'13"W
22 May 2010 19:00 Hdg 082.3 Spd 07.2 3°28'58"N 30°34'25"W

http://i46.tinypic.com/2jb7ddh.jpg

mm43

Ahhh. Light dawns on Marblehead. :) The latest positions show Seabed worker at search zone 2. I'm now sure the run up there on 5/22 was to drop and survey transponders, thus saving time when returning to search that area on 5/24.

The Stall Scenario
 

I suppose if AF447 did the decel into a deep stall scenario, it would be found much closer to the last known position than the bulk of the search has been looking at.
Looks like the doughnut hole inside of the present search area might be a prime candidate. After all, how did it hit so slow?? Had to be well under 200 knots just based on the condition of the bodies, and the large size of the found wreckage.
If AF447 was drifting heading slowly on its way down, it would be even closer to the LKP. Maybe they have been searching too far out?

Thanks for the SEABED WORKER position updates, mm43.

Auv-ee and Machinbird: But why search Zone 2 at the latitude where the first body was discovered, when it is known that the cluster of bodies in that area was moving northward at about 20nm/day? Am I missing some major eddy or whirl or stasis that would have kept the bodies neatly on station around 3º30' N from June 1 until June 6, when they then commenced their northward drift (see post #1120)? Got any ideas, mm43?

I do hope the strategic thinking and best-guesses behind the search scheme eventually see the light of day, and don't drown in increasingly difficult internecine politics (thanks for the sad Navy vs. BEA story from LE FIGARO, jcjeant). We interested outliers are certainly working from a great deal of obtuse and inferred data points, unlike the experts presently in the kitchen. As I've said before, it would be nice to have regular information-rich updates direct from the scene -- a facility easily arranged in this day and age. But then that information might be more difficult to manage.

GB

Yes my post was challenged very quickly. I was incorrect and I now agree fully with takata, in that beyond a certain depth, a body will not float to the surface.

Zone 2 was an area of uncertain coverage by the Orion towed sonar. Whether or not anyone, including BEA, actually believes the a/c could be in zone 2, I expect they need to be able to say that they covered the whole search area completely, and there is no possibility that the target was missed. Then, if there is a 4th attempt, they don't have to revisit or defend old uncertainties.

The surface current was a bit of an unknown quantity between 3°N and 3°30'N, though in theory it should have been NE and turning through to N at 3°30'N and then depending on the Long would take you to the north of TASIL or further to the west. The velocity would appear to have been a little less than that observed post 6 June, and knowing the areas searched by the FAB, that effectively means that the current was moving NNE through to N to the west of the LKP. It could well be that a stream was flowing from around 2°30'N 31°20'N in a NNE direction and on approaching 3°N branched both left to N and right to NE, then sort of paralleled up again to become N by E. What happened in the middle would be anyone's guess.

The question of leeway for the bodies needs to be addressed, but post 6 June the wind was generally light and from an easterly quarter. Some bodies may have floated initially, and others will have sunk to an equilibrium point, and depending on water temperature at that point will have eventually made their way to the surface. Warm temps mean this happened more quickly than in temperate latitudes.

The Port Outer Spoiler was found 40NM N by E of TASIL on 13 June, and if the bodies and the vertical stabilizer hadn't been subject to some leeway due to windage, I would have expected them all to have finished up passing close by TASIL. That wasn't the case, and the vertical stabilizer was located some distance to the west of the nearest bodies, which indicates that the wind had played a part in its total drift.

My first look at the problem using OSCAR and QuikSCAT data lead me to place the impact position near 3°N and 31°10'W, and one of those positions was south of the LKP and hasn't been looked at in this phase 3 search. My present thinking based on updated information still comes up with a position near the north side of the southwest quadrant at around 2°49'N 31°06'W. In due course I'll put up a graphic to show how that is achieved, but in any-case the fact that most of the area to the north has been searched and found devoid of any debris, doesn't mean that my position is any better that any other guess at this stage.

I'm disappointed to read that according to Le Figaro the BEA and the Ministry of Defence are having a spat over the existence or not of pinger records.

mm43

Hello mm43,
Opinions of experts about it is that bodies which will sunk will sunk directly to the bottom of the sea while bodies which will float will be almost invisible (99.9% submerged) from the surface during the first few days (which explain why they could have been flown over at close range without being spotted during the initial aerial search and rescue phase). There is no equilibrum point to be reached at any depth (passed 10 meters the pressure is twice the surface).
I'm very disappointed about that too. Most of this clash is due to hastily communication from MoD (not from Navy) which disrupted BEA's systematic search plan while those data aquired were still too raw to be fully exploited.

The good point is that Emeraude's sensors actually recorded those pingers along 20 positions and that they are actually very real. As it is explained, the uncertaincy about location is still great because distances are very difficult to compute due to the complexity of the relief. It means that the search zone may be far away than first computed. The example given about the localisation of Lebanon pingers +10 nautic miles farther than supposed range of emission is a good clue about all the difficulties encountered to pin point an exact location.
Now, the positive point is that a fourth search will be conducted with certaincy and that only time will permit to fully analyse Navy findings.
S~
Olivier

Body buoyancy and drift
 

takata and FlexibleResponse,

Thanks for revisiting your discussions of 11 months ago, some of which I had missed. Am looking forward to mm43’s promised graphic-explanation of his estimated impact positions, which have been consistently well to the west of the body-based one that you (takata) proposed last June 21st. Are body reverse-drift calculations more reliable than debris, due to less windage-effect?

Chris

Tape mixup?
 

There are some puzzling details in Fabrice Amadeo’s FIGARO story.
On 6 may 2010 the website of french weekly ‘Le nouvel observateur’ quotes general Christian Babtiste, spokesman for the ministry of defense as saying that the pinger sound has been identified on tapes obtained on 1st july 2009:
From BEA 2nd Interim Report, Appendix 2:
My translation:
During the search operation, the SNA support base has conducted tests with a ship of the same type in the Mediterranean sea to verify and optimize the performance of the sensor that was used. The results of these trials permitted to define new settings to increase the detection capability of Emeraude’s receiver (detection range 2000 m from 10 to 30 june, extended to about 3200 m from 1st to 10 july).

In other words, the tapes obtained in the mediterranean were dated before july 1st.

HN39

Lost:The Mystery of Flight 447
BBC2 Sun 30 May 2010 22:00 &
Wed 2 Jun 2010 00:20


BBC - BBC Two Programmes - Lost: The Mystery of Flight 447

Has a more precise map been publicly published of where all the larger pieces of the plane were recovered, and when? For example, if I recall correctly, at least one of the cabin crew seats was recovered, presumably entire with harness and frame, as it was determined that the seat was unoccupied at the time of impact. (Which I found rather surprising by itself.) I suspect that enough of the seat would be submerged as to introduce a considerable drag and be less affected by wind-induced drift.