AF 447 Search to resume [Archive] - Page 13

Squawk_ident

5th Apr 2011, 09:32

interview of Jean-Paul Troadec, BEA manager.05APR2011

ouest-france.fr - Rio-Paris*: un mois pour remonter épave et corps (http://www.ouest-france.fr/actu/actuDet_-Rio-Paris-un-mois-pour-remonter-epave-et-corps-_3636-1752870_actu.Htm?xtor=RSS-4&utm_source=RSS_MVI_ouest-france&utm_medium=RSS&utm_campaign=RSS)

.../...

Q : Cette zone proche de la dernière position connue n'avait pas été explorée. Pourquoi ?
A:- Entre la dernière position connue et l'impact, il s'est écoulé cinq minutes. Le Rio-Paris aurait pu parcourir une distance très importante. Au départ, nous n'avions pas de raison de penser qu'il était pratiquement à l'aplomb de cette dernière position.

Q : This area near the last known position was not explored. Why ?
A: Between the last known position and the impact, five minutes elapsed. The Rio-Paris flight could have cover a very important distance. At the beginning we had no reason to think that he was almost straight above this last position.

Est-on certain qu'il n'y a pas eu explosion en vol, mais décrochage ?
- L'explosion paraît très peu vraisemblable. Les débris légers retrouvés à la surface indiquaient déjà, par leurs positions symétriques, que l'avion avait touché l'eau en ligne de vol et en étant entier. Il n'était pas préparé à l'amerrissage car les passagers n'étaient pas équipés de gilets de sauvetage. Quant à l'hypothèse d'un décrochage, il est trop tôt pour se prononcer.

Q : Are we sure that there was no in-flight explosion, but stall.
A : An explosion seems very unlikely. Light debris found at the surface already indicated, with their symmetrical positions, that the airplane had touched the water in line of flight and was entire. It was not prepared for a sea landing because passengers were not wearing life jacket. About the stall hypothesis, it is too early to give an opinion.

Remonterez-vous tous les corps des passagers ?
- L'identification des victimes, quand elle est possible, s'impose à nous. Certaines familles le souhaitent, d'autres pas. Mais l'on ne peut pas remonter les corps de façon sélective. Les familles vont être associées à cette ultime phase des recherches. Nous allons les réunir prochainement et leur détailler la façon dont cette mission va être conduite.

Q: Will you retrieve all passengers bodies.
A: Victims identification when possible is essential to us. Some families wish this, others no. But we cannot retrieve bodies in a selective manner. Families will be associated to this last searches phase. We will meet them together soon and detail them the way this mission will be managed.
.../...

The wreckage is at 3900 meters depth. Three boats are considered for this operation:
René Descartes owned by France Télécom Marine, based at Brest, France. Another based in Cyprus and a third owned by Alcatel Lucent. The selected ship will be on the scene in one month.

takata

5th Apr 2011, 10:44

A: Between the last known position and the impact, five minutes elapsed. The Rio-Paris flight could have cover a very important distance. At the beginning we had no reason to think that he was almost straight above this last position.

Here is the main point that is almost ruling out an upset at 02.10 causing the crash where the wreckage is actually lying.

It certainly doesn't take 5 minutes to "deep stall" from cruise level to sea level, when in the meantime, the aircraft is circling at the same spot and will keep this attitude at impact.
Something else happened in that sequence that we still don't know.
CVRs and FDRs are really the key to fully understand it.

I'll bet that they will recover the engines.

grity

5th Apr 2011, 10:46

I think, this are the windows in seat row 29...31 the first behind the emergencydoor, grity
http://www.bea.aero/fr/enquetes/vol.af.447/images/wreckage.jpg

Chris Scott

5th Apr 2011, 10:51

Interesting discussions about deep stalls, but do the revelations of the last 48 hours change anything? ;)

Machinbird,
For the benefit of this pedestrian pilot, what is a "cat shot", and are you talking about a F-102/ F-106 (delta-wing)?

Quote from SaturnV (post #3000):
"...there was no lightning with this particular complex; none detected by satellite, nor by ground stations. Apparently lightning is infrequent in the ITCZ."

Before anyone gets the idea that sub-Saharan Africans (for example) know little about lightning (!), my understanding and experience suggest ITCZ lightning is less likely over the oceans.

I'm interested in the idea that some serious atmospheric changes may have been experienced if they got far enough to emerge from the north face of that Cb. As I've said before, this would have left them very poorly placed if they were already trying to fly thrust/attitude with unreliable airspeed indications.

Chris

PS (Edit):
takata,
Maybe the five minutes involved a tear-drop?

SaturnV

5th Apr 2011, 11:04

From takata's list, here are the three recovery ships that are being considered. EDT ARES is presently off Alexandria Egypt.

Rene Descartes
RENE DESCARTES - 9248100 - Vessel's Details and Current Position (http://www.marinetraffic.com/ais/shipdetails.aspx?mmsi=226291000)

Île de Bréhat
ILE DE BREHAT - 9247053 - Vessel's Details and Current Position (http://www.marinetraffic.com/ais/shipdetails.aspx?MMSI=635230000)

EDT ARES
EDT ARES - 9130755 - Vessel's Details and Current Position (http://www.marinetraffic.com/ais/shipdetails.aspx?MMSI=212858000)

EDT ARES aircraft recoveries:

EDT and Phoenix Recover Yemenia Flight Debris (http://www.edttowage.com/en/news-7-EDT_and_Phoenix_Recover_Yemenia_Flight_Debris.html)

Aug 20, 2010 (http://www.phnx-international.us/Aug%2020,%202010.htm)

EDT Ares assists with search - Offshore Shipping Online (http://www.oilpubs.com/oso/article.asp?v1=4251)

ZeeDoktor

5th Apr 2011, 11:45

takata

5th Apr 2011, 12:07

@takata: Let's look at what we know.
- The last position received was at 02:10:34. That's 20 seconds after the AP switched off and problems obviously had begun.
- At 02:12:51, ADR disagree triggered, with concurrent loss of AOA protection

- Assuming they had taken evasive action shortly after 02:10:34 by initiating a left turn, they would now (~2.5 minutes later, at 02:12:51) be just about overhead where the wreckage was found and would probably be just about coming out of the left hand course reversal

- Enter a deep stall / flat spin at that point, facilitated by lack of AOA protection, pitch and power flying in turbulent conditions in a turn

- Assuming ~350ft/s descent rate in flat spin, impact with water would have happened ~1m40s later (if still at FL350 when entered)

- time predicted for impact: 02:14:31, that's 6 seconds off from when transmissions ceased. Very close.

My understanding is that the system switched to "Alternate law 2" from the begining at 02.10. Then, "AOA protection" was lost from the start of the sequence anyway.

Why would they initiate such an "evasive action" at 02.10 while they had to keep flying "pitch and thrust"?... this would not be very smart without a real need to do so. During all the similar sequences documented so far (more than 36), not a single crew did take such an initiative to make an "evasive action" only due to unreliable airspeed data.

Nothing is telling us for sure that the aircraft impacted at the end of the ACARs transmition (02.15). In fact, without engine power, no more ACARs would be sent anyway.

ZeeDoktor

5th Apr 2011, 12:31

chaz88z

5th Apr 2011, 12:35

At first, I would like to thank you all for this really interesting thread !

After reading your posts I was thinking of an unrecovered spin as there is not much distance between LKP and crash site.

But ZeeDoktor did the maths and found 350ft/s... Not likely to be a spin...
That makes it 21.000ft/min almost a 200kt vertical speed (free-fall?). What about the fact that BEA hypothesis is a close to horizontal water contact.

I have further questions some people following this thread might answer :

What are the consequences of a high speed stall ? On a swept back wing ?

My guess are : First you lose altitude then static pressure rise and compressability effects are weaker.
Assuming this, would the airfoil recover lift by itself after a high speed stall ?

What if only one wing enters a high speed stall ? Self destruction might be more likely than spin at that speed isn't it ?

takata

5th Apr 2011, 12:53

Only alpha floor is lost in alternate law 2. AOA is still monitored and protection now applies to stall speed rather than AOA. Protection is lost only if VSG1 can no longer be calculated (which I think was not the case, happens when slat/flap information is not available and GWT is no longer known), or when dual ADR fail or disagree is triggered.

In our case, it looks as if at least stall protection was available until ADR disagree triggered.

I agree evasive action in the form of a turnaround is virtually unheard of (in fact it's usually discouraged for reasons we can discuss in another thread). But we don't know what picture was painted on the radar. Nor what went on with the aircraft, maybe a something evolved that prompted them to return to the nearest landing opportunity which would have been behind them (I think?)

Well. but the late triggering of "ADR disagree" at 02.12 is obviously what caused ALT2, Autopilot OFF, Autothrust OFF, Rudder limiter control, etc. And this is what happend in the first place at 02.10 because of those Pitots reading errors... There is not so much possibility for the system to switch to ALT2 in any case.

The fact that this maintenance message was transmitted much later in the ACAR sequence doesn't mean that the system was not already in error before its maintenance time stamping or you'll have to find an explanation why ALT2 was already triggered at 02.10.

ZeeDoktor

5th Apr 2011, 13:15

@takata: Eh bien, you're right the timestamp of the ACARS message is not necessarily related to when the fault occurred. The sequence of events up to the ADR disagree message however would be conducive with a disabled pitot and the corresponding crew action such as manual AP OFF.

@chaz: You don't need that high a speed. The point is I don't know how fast a flat spinning A330 will descend. Any number between 7000 to 20000 ft/min sounds reasonable to me: If you assume FL350 to FL0 in 5 minutes, you just need 7000ft/min ROD. If you take the highest number I've seen (taken from someone elses calculations on the previous page, look for the diagram with velocity vectors), the numbers can still be plausible.

Hopefully we'll know for sure soon. I was just (for my own curiosity) trying to satisfy possible events between LKP and what we now know as the wreck position.

takata

5th Apr 2011, 13:26

@takata: Eh bien, you're right the timestamp of the ACARS message is not necessarily related to when the fault occurred. The sequence of events up to the ADR disagree message however would be conducive with a disabled pitot and the corresponding crew action such as manual AP OFF.

AP OFF was certainly not due to manual action: AP was switched OFF by the system, hence, this relevant ACAR was sent (no ACAR would be sent if this was due to a manual switch). It means that before this point, AP was still ON, and after this point, it was automatically switched OFF because ALT2 was triggered (hence, because ADR disagree due to pitot errors)...

BJ-ENG

5th Apr 2011, 13:48

Re Chaz: Pulkovo flight 612 stall. From FDR data; stall from 11,677m reached vertical velocity of 94m/s around 5730m, decreasing to 72m/sec just prior to impact. Time of decent was 126sec.

Re the engines. Unless I am very much mistaken, the LPC fan blades are gone, the front sections of the outer nozzle containment shroud ring has gone, and what we see in both pictures are the remnants of the guide vanes which form the rear part of the bypass nozzle. Surely, this level of destruction is suggestive of a high engine shaft rpm at impact.

Compare this with the crush damage sustained on engine 2 for TW 800 and the still intact nature of the LPC fan, where the NTSB concluded that there was minimal amount of low-pressure rotor rotation at impact.

General Electric cf6-80
ImageShack® - Online Photo and Video Hosting (http://img837.imageshack.us/i/96577608.jpg/)

ImageShack® - Online Photo and Video Hosting (http://img607.imageshack.us/i/generalelectriccf680.jpg/)

CROSS SECTION
ImageShack® - Online Photo and Video Hosting (http://img859.imageshack.us/i/cf680c2.jpg/)

FLIGHT INTERNATIONAL: DETAILED SECTION
ImageShack® - Online Photo and Video Hosting (http://img546.imageshack.us/i/generalelectriccf680c2.jpg/)

REAR VIEW OF FAN SECTION
ImageShack® - Online Photo and Video Hosting (http://img833.imageshack.us/i/80525171.jpg/)

TW800 ENGINE 2 - IMPACT CRUSH DAMAGE
ImageShack® - Online Photo and Video Hosting (http://img851.imageshack.us/i/no2twa800.jpg/)

ACLS65

5th Apr 2011, 14:18

@Chris Scot I believe he means the catapult launch from an aircraft carrier.

Chris Scot
For the benefit of this pedestrian pilot, what is a "cat shot", and are you talking about a F-102/ F-106 (delta-wing)?

jcarlosgon

5th Apr 2011, 14:32

By takata:

"Well. but the late triggering of "ADR disagree" at 02.12 is obviously what caused ALT2, Autopilot OFF, Autothrust OFF, Rudder limiter control, etc. And this is what happend in the first place at 02.10 because of those Pitots reading errors... There is not so much possibility for the system to switch to ALT2 in any case."

Rudder Limiter. Fast Vertical Speed fall. Did the vertical stab last untill impact?

lomapaseo

5th Apr 2011, 14:33

I don't agree with the post above engine assessment.

Suffice it to say that the high quality photos available to the investigators would be enough for them to make a determination. The photos we are looking at do not have resolution to assess the blade condition

The TWA engines came off during the spin to the ground and thus would be no higher in RPM than a very low windmill (they fall flat).

My interest is the ten or so seconds of flight before the error codes were set and not at what the engines looked like when they hit the water.

Turbine D

5th Apr 2011, 14:54

I think the only things you can conclude from the photos of the one engine are these:

1. The fan rotor and fan casing are both gone.
2. Most of, if not all of, the fan bypass stator vanes are gone.
3. The titanium fan frame and struts are visable but damaged.
4. The one piece turbine rear frame containing the rear engine mounts (attachment point to the pylon) is gone as well as the rear portion of the LPT casing exposing some turbine rotor blades.

TD

takata

5th Apr 2011, 15:03

Rudder Limiter. Fast Vertical Speed fall. Did the vertical stab last untill impact?
So far, all the rudder damages analyzed were fully consistent with an impact as it is described by the investigators findings. But one is free to believe whatever theory he likes more than any stuctural analysis (instead of checking all the published reports about it).

The structural analysis is telling that the damages were mostly due to vertical compression all over the airframe (and symetrical) with secondary horizontal forces and very few rotational forces. It doesn't mean "Fast vertical speed fall" but that the vertical pressure was higher than horizontal forces.

This is consistent with a stalled level slightly pitched up airframe at impact.

Now, what caused this end stall?
Did this aircraft departed from controled flight at cruise level or did something else happened in between that we are still not aware of?

ushumgal

5th Apr 2011, 15:23

takata

5th Apr 2011, 15:57

Now, something else than the "deep stall theory" should not be ruled out at this point.

Even if unpowered, such an airframe can glide very far (more than 120 NM from FL350) when it is able to trade off altitude vs distance. It would defintively not fall from the sky like that if it could remain wings level, with all its thrust or if it had enough altitude.

It would take a fair amount of structural damages or an incapacited crew to go down like that (like that?). More likely, before impact, she was left at low altitude without thrust and may have unexpectedly stalled at some point (hence no preparation for ditching).

There was already in the past some problems related with those engines (CF6-80E1 in this case CF6-80E1A3) ice/water ingestion:

FROM : AIRBUS CUSTOMER SERVICES TOULOUSE
TO : ALL A330 OPERATORS
OPERATORS INFORMATION TELEX - OPERATORS INFORMATION TELEX
AND
FLIGHT OPERATIONS TELEX - FLIGHT OPERATIONS TELEX

TO: ALL A330 OPERATORS
SUBJECT: ATA 72 - A330 DUAL ENGINE FLAME OUT
OUR REF: SE 999.0069/JS dated 09 JUNE 2006
CLASSIFICATION: INCIDENT - ADVICE (FLIGHT OPERATIONS)
REFERENCE
- OIT SE 999.0067/06/JS dated 02 JUNE 2006

1. PURPOSE

The purpose of OIT/FOT is to provide latest update and to provide operational recommendations on the dual engine flame out event reported through OIT ref. SE 999.0067/06/JS dated 02 JUNE 2006.

2. EVENT DESCRIPTION

On 1 June 2006 an A330-200 aircraft, equipped with General Electric (GE) CF6-80E1 engines, experienced a dual engine flame out during descent around flight level 200.
Both engines quickly recovered and a safe landing was performed.

3. INVESTIGATION STATUS

The investigation into this event is led by the Investigation Authorities with assistance from Airbus. The investigation is still in its early stages, however, DFDR preliminary analysis has shown that:
- During descent, while engines started to accelerate for aircraft altitude capture, both engines flamed out simultaneously.
- Both engines automatically relit after flame out, and recovered within approximately 45 seconds.
- Engine Anti Ice had been selected ON during the descent, and Wing Anti Ice had been selected ON shortly prior to the event.
- Aircraft systems behavior was normal including automatic RAT extension.

Boroscope inspections have been performed on both engines without significant findings.

Based on the above, the initial Airbus/GE view is that this event is similar to other power loss events at altitudes above 10 000 ft attributed to inclement weather as experienced on CF6-80 engines installed on various aircraft types.

The aircraft returned to service on 7 June 06.

4. OPERATIONAL RECOMMENDATIONS

Waiting for final investigation results, in order to increase the fuel/air ratio in the engine so as to mitigate the possibility of experiencing an engine flame out, the following provisional procedure is recommended:

. If inclement weather/icing conditions are expected at any time during descent or if convective activity is identified by the weather radar in the vicinity of the aircraft flight path:

At top of descent:
- ENG ANTI ICE__ON
- WING ANTI ICE_..ON
- PACK FLOW___HI

Below 10000 feet :
Resume normal anti ice and pack flow operation according to weather conditions.

Note that the fuel consumption and the idle thrust will slightly increase when selecting ENG ANTI ICE_ON, WING ANTI ICE_ON, and PACK FLOW_HI.

Final operational recommendations will be implemented in the FCOM/QRH via OEB or TR.

5. FOLLOW-UP PLAN
An update will be provided by 16 June 06.

eugenefraxby

5th Apr 2011, 16:04

AirculePoirot

5th Apr 2011, 16:05

Hello everyone – first time here - always wanted to be a pilot, but became an electronic engineer, then a teacher.

This is a fascinating thread, and I've learned a huge amount, especially about stalls in the last few pages. I have a question for the experts:

I recently saw a program on TV that explained the phenomenon of super cooled water vapour which can exist in very pure air, in the absence of particles to condense around. The program narrator shower how ice spontaneously form around an object (he used a real pitot tube) when introduced to supercooled water vapour in the lab

If this did happen, it would be more than a pitot tube or three that iced up, especially the wings and the leading components of the entire craft. (Some manner of) stalling would have occurred, but the subsequent loss of altitude would have cause the ice build-up to melt, and the possibility would exist for the aircraft to become manageable again, would it not?

What are the chances that the plane crashed in a belated but vain attempt by the pilots to pull out of a dive? Is it true that ACARS transmits error conditions, but not the reversal of a error condition back to OK?

takata

5th Apr 2011, 16:22

@takata - You're clearly right that if the aircraft can maintain wings level at a suitable attitude, it can glide for a long time from FL350.

What if the aircraft entered a storm cell? The huge updraft/downdraft velocities inside the storm system could make the level aircraft lose altitude very rapidly?

I only fly gliders, not large aircraft, but I've flown in pretty rough conditions where we were wings level at normal attitude experiencing 10kt downdrafts (which is a very big downdraft for a glider at little more than 1000ft!). Under those conditions you can lose altitude very rapidly even if the aircraft is in a normal, non-upset condition.

Right, but we still don't know yet what were precisely the real weather conditions during this flight. As I said above, something still unknow happened between FL350 and the aircraft final attitude before impact at sea level. All we know for sure is that it took at least 5 minutes (maybe more) to go down and that the weather (ice, rain, storm) was certainly the primary factor above anything else.

All those pitot tubes don't freeze in normal/severe weather conditions, neither a dual engine flameout would be caused, neither such an aircraft would be lost with all hands from its cruise level.

This is why we should learn much more from the recorders (if they are found) and the wreckage analysis (including those engines remants).

BJ-ENG

5th Apr 2011, 16:26

@Lomapaseo

Sorry, missed your earlier post about non visible blades. Missing fan assembly and casing could also be explained by shearing forces as you suggest, rather than high rotational energy disintegration.

As you say, let's wait for better photos.

wes_wall

5th Apr 2011, 16:33

takata

I doubt very little controlled gliding was done during the descent. Rather I suspect that immediately following the onset of their upset, the airplane was quick to begin a loss of altitude, with a corresponding increase in vertical speed as forward ground speed declined. The apparent fact that flight attendant stations (seats) were not occupied (on impact) and the lack of any kind of cabin preparedness gives credibility to the quickness of events and the violence of the upset .

grity

5th Apr 2011, 16:56

http://d.yimg.com/a/p/ap/20110404/capt.f8cf432916b647c38f3b162f3933919d-f8cf432916b647c38f3b162f3933919d-0.jpg?x=400&y=282&q=85&sig=61CEi1P_8jxmh7N33euD5Q--
what can be the struktur right forward of the wing?
part of the wing or part of the fuselage? grity

takata

5th Apr 2011, 16:56

I doubt very little controlled gliding was done during the descent. Rather I suspect that immediately following the onset of their upset, the airplane was quick to begin a loss of altitude, with a corresponding increase in vertical speed as forward ground speed declined. The apparent fact that flight attendant stations (seats) were not occupied (on impact) and the lack of any kind of cabin preparedness gives credibility to the quickness of events and the violence of the upset.

Sure. But the real question is when did this final upset (causing the final crash, possibly stalled) occured?
At 02.15, 02.16, 02.17? Nobody knows actually.
Just before 02.15, at least one engine was obviously still running (certainly both) as ACARS were still transmitted...
Now, at impact time, this aircraft seems to be controlable in the case she was still powered. Was she?
How could we know for sure that she impacted at this exact time if she was not powered anymore?
I'm just pointing that this airframe is able of gliding unpowered but will need some altitude to do so, as well as being able to build up some speed to restart its engines.
When short of both, she certainly will go down.

BOAC

5th Apr 2011, 17:00

Come on everyone - we have VERY LITTLE more information now than we had 3000 posts ago. Let's focus on the 'search' and now 'recovery' rather than another 3000 'maybes'. I don't think I could face wading through another 3000 flat spins/stalls/ glides/turnbacks etc etc while trying to find mm's gems.

Personally I suspect the tail section (and FDR) are sadly a long way away from this wreckage - I hope I am wrong.

HazelNuts39

5th Apr 2011, 17:00

Only alpha floor is lost in alternate law 2. Maybe I'm just getting confused, which wouldn't be the first time. However, my understanding is that alternate law does not maintain any of the protections in pitch, except maneuver protection. (ref. FCOM 3.04.27 P6). The normal "High angle of attack protection" is replaced by a new protection called "Low speed stability". (ref. ACA Memo P6). An aural "STALL, STALL, Stall" warning sounds at low speeds, but may also sound at high altitude where it warns that the aircraft is approaching the angle of attack for the onset of buffet. (ref. FCOM 3.04.27 P6). BEA report #2 explains that stall warning is triggered when the AoA exceeds a certain threshold that varies with Mach number. Low speed stability is described in FCOM 1.27.30 P2: Alt 1: At low speed, a nose down demand is introduced in reference to IAS, instead of angle of attack, and alternate law changes to direct law. It is available, whatever the slats/flaps configuration, and is active from about 5 knots up to about 10 knots above the stall warning speed, depending on the aircraft's weight and slat/flaps configuration. A gentle progressive nose down signal is introduced, which tends to keep the speed from falling below these values. In addition, audio stall warning (crickets + "STALL" synthetic voice message) is activated at an appropriate margin from the stall condition. The PFD speed scale is modified to show a black/red barber pole below stall warning. Vα prot and Vα max are replaced by Vsw (stall warning speed). The α floor protection is inoperative.

Alt 2 is identical to protections in Alt 1, except that:
1. There is no bank angle protection in Alt 2 law.
2. In case of failure of 2 ADRs, there is no low speed stability.
3. In case of failure of 3 ADRs, there is no high speed stability.My interest in collecting these quotes is just to get this complicated subject clear in my own mind. My curiosity concerns in particular the first sentence of the Low Speed Stability description: "At low speed, a nose down demand is introduced in reference to IAS, instead of angle of attack, ...". Does this apply at high altitude? What if the reference IAS is erroneous, i.e. 2 ADRs providing similar low airspeeds? Was the gear possibly down and locked before impact?

takata

5th Apr 2011, 17:13

Personally I suspect the tail section (and FDR) are sadly a long way away from this wreckage - I hope I am wrong.
This hypothesis would be known quite fast as they have said that it would take only a few days to complete the whole reconnaissance and geo-localization of the debris field.

If the tail separated from the airframe before impact, why was the cabin still pressurized? Do you mean that it happened at low altitude or that it separated without releasing the oxygen masks or without compromising the cabin pressure?
If so, it would not be that far from the main debris field anyway.

snowfalcon2

5th Apr 2011, 17:34

However,
I'd imagine that even if the tail and FDR would have fallen off in the air, the pieces would certainly not glide very far (meaning more than a mile or so from the main part).

Fingers crossed for next news...

Chris Scott

5th Apr 2011, 17:42

lomapaseo

5th Apr 2011, 17:45

Personally I suspect the tail section (and FDR) are sadly a long way away from this wreckage - I hope I am wrong

Good point for a little discussion to while away the boredom

Given that we accept a flat impact with the water (liitle roll or pitch) would that give a clue whether a tail (possibly with recorders) was still attached ?

BOAC

5th Apr 2011, 18:04

No, nothing to do with in-flight separation 'theories', but more to do with a heavy impact, surmised to be tail down. It is expecting a lot of a tailplane assembly in that situation not to cause tail section separation which leaves a light'ish structure to make its own way down to the sea bed, taking with it the FDR. No pinger any more - now find it?

As I say, I hope for all I am wrong. The sonar 'finding' so far has been brilliant. I hope it continues.

SaturnV

5th Apr 2011, 18:45

BOAC, the distance between the location and the western boundary of the June 1 2009 search grid flown by Brazil is probably several nautical miles. If the VS came off simultaneous with the rest of the tail, I would think the VS would be hard to miss if the separation occurred within that first day's search area.

So I rather doubt that the tail is a long way off; however, it may be the most shattered section in the whole debris field. A lot of photography ahead.
____________

Below is this excerpt from the Canadian investigation into AI 182, which may be relevant to AF 447 if the tail did shatter.

“The depth of the wreckage varies between about 6000 and 7000 feet, and the effect of the ocean current, tides and the way objects may have descended to the sea bed was not determined, thus some distortion of an object's relationship from time of water entry to its location on the bottom cannot be discounted. In general, the items found east of long 12∞43.00'W are small, lightweight and often made of a structure which traps air. These items may have taken considerable time to sink and may have moved horizontally in sea currents before settling on the bottom. Marks left on the sea bed beside some wreckage does indicate horizontal movement of the wreckage as it settled.” "

http://www.publicsafety.gc.ca/prg/ns/airs/_fl/CASBai-en.pdf

I might add the entire wreckage trail for AI 182 was 6.5 NM, on a line 110 degrees true.

RatherBeFlying

5th Apr 2011, 18:49

I suspect that WHOI and/or the BEA people onboard know exactly where in the structure the CVR/FDR are located along with what they look like and will be giving extra scrutiny to the survey photos of the rear fuselage debris as they are retrieved.

I earlier suggested placing the CVR/FDR in the fin as it seems a buoyant and easily retrieved structure.

robertbartsch

5th Apr 2011, 18:53

Can someone comment on the recent CNN report that said oxygen masks were not down.

I would think this would indicate positive cabin pressure and no structural damage. How about the recovered stabilizer; what does that tell us?

mm43

5th Apr 2011, 19:25

I would think this would indicate positive cabin pressure and no structural damage. How about the recovered stabilizer; what does that tell us?There was nothing in the ACARS messages to indicate that cabin pressure had been lost, but rather to the contrary, i.e. the aircraft was descending at a greater rate than the cabin pressure controller could cope with. On impact the cabin pressure will have been less than the outside pressure, which would have caused a small implosion in conjunction with the major impact implosion.

The Vertical Stabilizer told us that it had left the empennage as a result of the tail impacting with the sea surface, as both compression and tension damage was observed.

bearfoil

5th Apr 2011, 19:28

There is no reason to believe the VS separation (If at altitude) included the HS, elevators, or for that matter compromised the integrity of the aft bulkhead. So pressurization may well have existed at impact.

I for one agree with BOAC, as those who are vets of the prior threads know. I have believed in VS separation from the beginning (perhaps as early as initial upset). The location in the initial search of the port elevator suggests a separation also of the Tail feathers, but again, not necessarily at impact.

The early conclusions of BEA notwithstanding, to entertain that the VS was aboard at final impact challenges logic, as the condition of the sea floor debris would suggest. The galley stack included, the VS and cabin parts were no where near the final impact zone, as the extent of the destruction suggests, from the new photos. It also requires a substantial horizontal ("Thrown forward" and clear of) component, unlikely to have happened even in consideration of BEA's conclusion.

The condition of the VS has been exhaustively discussed prior, and this is not an invitation to continued speculation, at least by me.

Machaca

5th Apr 2011, 19:36

The DFDR is mounted just aft of the rear pressure bulkhead:

http://i337.photobucket.com/albums/n385/motidog/A330aftBHtail1.jpg

mm43

5th Apr 2011, 19:42

Machaca

Another one of your good pics. Thank you.

I believe the CVR is located about 2 frames aft of the rear door.

glad rag

5th Apr 2011, 19:59

Another one of your good pics. Thank you.Yep :D:D

Would that be FTI kit being installed?? I ask because of the orange loom you can see between the top ladder rungs........I'm not sure on the 330 but I thought the recorders would be in a pressurised area rack...could be wrong of course.

... ain't this thread so very interesting? here's hoping that the authorities can finally find some more info and the families some closure.

GR.

Turbine D

5th Apr 2011, 20:12

Here is a FAA presentation (Powerpoint) regarding high altitude operations and conditions, which encountered, could lead to upset and a stall condition that takes some quick recognition and controlled responses to recover from.

http://www.faa.gov/other_visit/aviation_industry/airline_operators/training/media/Appendix%203-E_HighAltOperations.ppt

TD

ACLS65

5th Apr 2011, 20:37

Turbine D's Powerpoint opens fine in OpenOffice which is free to download if you need it.

OpenOffice.org Downloads (http://download.openoffice.org/)

chris weston

5th Apr 2011, 21:13

Opens fine in Google Chrome.

CW

Lonewolf_50

5th Apr 2011, 21:27

Thanks for the FAA brief, it was very informative for those who don't fly the big iron way up high on a daily basis. :ok:

D Bru

5th Apr 2011, 21:51

I'm tempted to believe that, despite all "expert" interpretation and background delivered to BEA before recently, in the end the arguments delivered on this forum on a "deep stall scenario" have led to start out the 4th phase from the area close to LKP.

I see no other explanation, other than BEA's staunch belief from the onset that AF447 could not have gone down south of LKP and pretty much all of the area north of LKP had been searched in previous phases, except where the A/C has been found now.

Of course a systemic search would lead to results at soime point, but for once let us applaud BEA, Airbus and AF that they got it right at this point.

That bypasses (and quite rightly so) a number of interested or fascinated people, including myself, who were uptil now convinced that the A/C could not be near LKP.

With the location of the debris field there is now a 100% probability that the FDR and CVR will be retrieved and we'll finally know what happened.

Not only will we have two years after the events conclusive evidence to give those berieved an explanation of the cause, but hopefully also the definite clues of how to mitigate a similar catastrophy in the future.

It has been and it is clear to me that neither Airbus, AF or the french government are ducking their responsabilities.

Dutch

DjerbaDevil

5th Apr 2011, 23:08

The exact location of the AF Airbus 330 crash site has not been mentioned before but the following article in the Guardian indicates a crash site only a few hundred yards from LKP:

http://www.guardian.co.uk/world/2011/apr/04/air-france-plane-crash-victims-found?CMP=EMCGT_050411 (mhtml:{25171FF0-2B93-487E-847D-B68C941D8898}mid://00002515/!x-usc:http://www.guardian.co.uk/world/2011/apr/04/air-france-plane-crash-victims-found?CMP=EMCGT_050411)&
"Scanning a circle with a circumference of around 75km from where the plane is thought to have crashed, the robots sent pictures back of wreckage on the seabed a few hundred metres west of the last known position of the plane."

Machinbird

5th Apr 2011, 23:33

the following article in the Guardian indicates a crash site only a few hundred yards from LKP:I'd take that as a reporters interpretation of something he may have mis-heard.
Without information very close to the BEA as the source, I would be very skeptical of anything I read in the papers.

SaturnV

5th Apr 2011, 23:48

DjerbaDevil. the Guardian report is in error. The search at this point was not scanning a circumference that was approximately 40+ NM from the LKP. So the several hundred meters from the LKP is also likely wrong.

This journalist couldn't even get the flight destination correct in the initial article; he or she had the flight going from Paris to Rio.

promani

5th Apr 2011, 23:57

Many years ago I was taught that in any search exercise, you always start at the LKP and work outwards from there, allowing of course for any info that may direct the search in a particular direction. If this had been done last year, then maybe we would now know what really happened that fateful morning. It seems that it took the Russians to enlighten the French on a likely crash site. Also we can exclude the oil slick now, as I hinted awhile ago?

Thermaller

6th Apr 2011, 00:40

Considering that the released images appear to have been carefully selected so as not to show the remains of the deceased and not to reveal details that may either explain or lead to speculation about the cause of the crash, we should consider that the rear of the aircraft is in fact in the debris field and has been located.

Does anyone have knowledge of the French legal system? WHOI is contracted to BEA so presumably the recovered boxes will be handed to the BEA. Let's hope there are proper procedures in place on the recovery ship. Since there are criminal charges involved, would the French court order that the FDR and CVR be examined under court supervision, or will the BEA have exclusive control over what happens to them.

Machinbird

6th Apr 2011, 01:19

Also we can exclude the oil slick now, as I hinted awhile ago? Promani, how did you reach that conclusion?

Thermaller, Alucia/WHOI, having just located the wreckage are mapping and documenting the site. The actual recovery will be by others under control of the French Government.

CONF iture

6th Apr 2011, 01:52

What if the reference IAS is erroneous, i.e. 2 ADRs providing similar low airspeeds?
If 2 ADRs tell the same lie, the flight law will not degrade to ALT but remain to NORMAL. The low speed stability would not be part of the game, but the A\THR would try to compensate for the erroneous low IAS ... to MMO or beyond. I think.

auv-ee

6th Apr 2011, 02:12

Many years ago I was taught that in any search exercise, you always start at the LKP and work outwards from there, allowing of course for any info that may direct the search in a particular direction.

Seems to me that is exactly what was done. 15,000 km^2 is a lot of area to search, so BEA was very interested in any way to reduce that area (for Phase 3). The drift study seemed like a plausible way to do that, so they arranged for some people with drift experience to work on it. The result was the search box for Phase 3. In hindsight, the lack of detection in the search box made it likely (now clear) that the drift study was wrong (currents at this site are unpredictable), but at the time there were people who agreed with the decision to use this data to focus the initial search, while others believed (correctly, if confirmed by the FDR) that an upset A/C could never have made it that far from LKP.

If this had been done last year, then maybe we would now know what really happened that fateful morning. It seems that it took the Russians to enlighten the French on a likely crash site.

Hindsight is 20/20. I'm not sure how much influence the Russian experience had on the Phase 4 search. My understanding is that a brute force search of the remaining 10,000 km^2 was planned, starting with the part within 20nm of LKP. The rational for the 20nm is contained in the Metron analysis, and is based on the sort of aerodynamic considerations that have been much discussed here. I am sure that some will argue that the aerodynamics should have been given more weight at Phase 3.

http://www.bea.aero/fr/enquetes/vol.af.447/metron.search.analysis.pdf

Also we can exclude the oil slick now, as I hinted awhile ago?

In the drift analysis, the pollution spot is discussed on pages 134-5. Here they explain how they backtracked the spot. Oil slicks are known to respond more strongly to wind than current, because they have no projection below a very thin surface layer that is wind driven. The report states that oil typically moves at 3-4% of wind speed, and that the wind was strong from the north. The backtracked position, based on wind and current, displayed in figure 27 (from yellow dot, back to orange dot), was based on down grading the motion to 2% of wind, as discussed on page 135. If they had stuck with 3-4%, I think the result would have been very close to the now known debris location. So, it's probably premature to say that the spot is unrelated.

http://www.bea.aero/en/enquetes/flight.af.447/phase3.search.zone.determination.working.group.report.pdf

Machinbird

6th Apr 2011, 02:25

Machaca, That is a great picture. Lets use it for discussion.:D

http://i337.photobucket.com/albums/n385/motidog/A330aftBHtail1.jpg

Didn't an A340 lose its DFDR on a tail strike a couple of years ago? A recorder in this position may well be torn loose during an impact such as AF447 sustained. I would expect a skin strength discontinuity in the transition from the pressurized cabin to the unpressurized tail. Failure of the skin just aft of the pressure bulkhead is then quite possible, even with the plentiful stringers visible. The DFDR and some associated structure may well be lying by itself, separate from the wreckage.
The heavy vertical structure aft of the DFDR appears to be a support for the THS actuator. How does it tie in to the structure above? An item like this would be forced upward like a piston by the impact from below and could well account for how the VS was thrown. It appears to be pointed at the aft VS support structure. The forward VS support structure is above and forward of the aft bulkhead if I recall correctly
A very helpful view of an area most of us never get to see. :ok:

promani

6th Apr 2011, 03:04

Machinbird/auv-ee, the reason for my comment is that I understood that the BEA had consulted the Russians, who told the BEA that they normally found aircraft wreckage very near the LKP when searching for a crash site.
I have to admit that I find it difficult to accept that the oil slck had travelled south of the LKP, but the debris and bodies recovered in June 2009 went north. But I am willing to accept your expert opinions. I am just posting my thoughts, guys.

Bobman84

6th Apr 2011, 04:37

Can we please keep this thread about the search and recovery only. Start a new thread if you want to discuss stalls etc.

The debris is dispersed over "quite a compact area" of about 600 meters by 200 meters (1,960 feet by about 650 feet), he said.

All the wreckage will be brought to the surface and sent to France for study, said Jean-Paul Troadec, head of the French air accident investigation agency, the Bureau d'Enquêtes et d'Analyses, or BEA.

"We want to know what happened in this accident, most particularly so it never happens again," he said.

Three companies bidding to raise the wreck have until Thursday afternoon to submit proposals, he said.

The operation should take three weeks to a month, and will be paid for by the French government at an estimated cost of 5 million euros ($7.1 million), he said.

PJ2

6th Apr 2011, 05:02

Machinbird;

Yes, an Emirates A340-500 had a tailstrike at Melbourne. The preliminary report is here (http://www.atsb.gov.au/publications/investigation_reports/2009/AAIR/pdf/AO2009012_Prelim.pdf). The DFDR broke loose due to failed latches when the tail struck the runway (after a weight > speeds miscalculation). The report has a photograph of the DFDR mounting box and the DFDR, which fell rearwards, through an opening and below to the floor, (lower skin). It always remained with the aircraft.

Bobman84;
The operation should take three weeks to a month, and will be paid for by the French government at an estimated cost of 5 million euros ($7.1 million), he said.
To place in some perspective the costs of searches and estimated recovery, over the four years it took to complete the recovery and investigation of the Swissair 111 MD11 accident at Peggy's Cove, Halifax, the Canadian TSB spent approximately CAD$57m.

PJ2

Machinbird

6th Apr 2011, 06:39

Thank you for that clarification PJ2. The DFDR of that A340 ended up in an opening directly below the THS actuator, below the "floor" of the compartment.

That was a 'love tap' compared to the bash AF447 seems to have taken.
Unless the DFDR ended up wrapped in other structure, I would expect it is free of its mount simply in view of the forces involved and crash attitude.
That tail compartment is likely torn open.
Hopefully that bit of oarnge will not be too heavily silted or otherwise covered if it is lying free.
Then a good photo interpreter will have a fighting chance of picking it out.

jcjeant

6th Apr 2011, 06:42

Hi,

would the French court order that the FDR and CVR be examined under court supervision, or will the BEA have exclusive control over what happens to them.

A representative of the law will (should) be there to put seals on the black boxes

mm43

6th Apr 2011, 08:05

Hopefully that bit of orange will not be too heavily silted or otherwise covered if it is lying free.
Then a good photo interpreter will have a fighting chance of picking it out.There is enough steel in the DFDR that someone out searching for gold using a $50 inductive loop detector would find it 600mm (2 ft) under.:E

In all seriousness, the technique may need to be used if the the orange brick has gone and buried itself.:\

Substitute the person for a towed magnetic anomaly detection (MAD) device.:ok:

vanHorck

6th Apr 2011, 08:56

mm43,

Welcome back to you! You have been dearly missed!

PFR

6th Apr 2011, 09:09

Firstly....
Dutch wrote:-
"It has been and it is clear to me that neither Airbus, AF or the french government are ducking their responsibilities".Would agree Dutch. Lets give some credit where credit is due - there are a lot of professionals engaged in this.

Secondly...
As has been touched on in a number of earlier posts, including that running in Rumours & News, "AF447 wreckage found" (hopefully the moderators will merge the two) what of the significance of the Landing Gear being extended - as appears apparent in the photo's so far released.
Could that be a crew action in some effort to arrest speed/gain control of the descent? Or is it a reasonable post impact result?

sensor_validation

6th Apr 2011, 09:09

My understanding is that the system switched to "Alternate law 2" from the begining at 02.10. Then, "AOA protection" was lost from the start of the sequence anyway.

Why would they initiate such an "evasive action" at 02.10 while they had to keep flying "pitch and thrust"?... this would not be very smart without a real need to do so. During all the similar sequences documented so far (more than 36), not a single crew did take such an initiative to make an "evasive action" only due to unreliable airspeed data.

Nothing is telling us for sure that the aircraft impacted at the end of the ACARs transmition (02.15). In fact, without engine power, no more ACARs would be sent anyway.

Given the approx location of the debris, is the fact that the 02:10:30 (assume this is exact time of position acquisition, transmission 4s later) location was "slighty west 3NM" of track INTOL - TASIL now more significant.

Yes was discussed at length previously, and can only be answered for sure by the black boxes - but was the deviation deliberate or indicative of an earlier upset?

HazelNuts39

6th Apr 2011, 10:23

what of the significance of the Landing Gear being extendedFor those still interested in the causes of this accident (and we're still months away from the data we're hoping for), yes, I think it would be very significant if the gear had been extended by crew action. It is of course quite possible that the gear was released from its uplocks by impact forces. However, I would think that the duration of high vertical acceleration is no more than a small fraction of a second, and therefore consider it unlikely that the gear would reach the fully down position before being ripped out together with its back-up structure in the wing.

Perhaps someone could produce a picture of that gear as installed on an intact airplane?

Regards,
HN39

takata

6th Apr 2011, 10:40

All the wreckage will be brought to the surface and sent to France for study, said Jean-Paul Troadec, head of the French air accident investigation agency, the Bureau d'Enquêtes et d'Analyses, or BEA.

Sorry, but it is not what Jean-Paul Troadec said. This is a bad translation of "All the wreckage brought to the surface will be sent to France", which is quite different: they will only brought to the surface what part is needed in order to fully understand the crash; they won't need the full aircraft if the recorders are recovered and are telling enough about the cause.

Shadoko

6th Apr 2011, 10:49

Hi,

First, thanks to all people giving so much interesting datas on this forum.

Second, as a newbie, three questions I did not find answers by reading some threads (too quicly?):

- if an engine flameout happened, was an ACARS to be transmitted?
- is a spinning "posture" making a stop to the satellite transmission (so, no more ACARS)?
- in the event of a vertical stabilizer rupture, is the cabin depressurization mandatory (and followed by an ACARS)?

Third, ... sorry for my frenchie English ...

Regards,
Sh

hetfield

6th Apr 2011, 10:50

I think it would be very significant if the gear had been extended by crew action. Isn't there a blocking valve in the hydraulic system to prevent gear extention above a certain speed?

SaturnV

6th Apr 2011, 10:53

The following figures are from the Metron analysis that was done for BEA in January 2011.

http://i136.photobucket.com/albums/q199/tahitiplage/SNAG-0169.jpg

^^^^ June 1 search grids. (low level only shown.) Brazil in yellow, France in blue.

http://i136.photobucket.com/albums/q199/tahitiplage/SNAG-0170.jpg

^^^^ June 3 search grids.

http://i136.photobucket.com/albums/q199/tahitiplage/SNAG-0171.jpg

^^^^ June 6 search grids. Voila.

Figures for the search grids flown on June 2, 4, and 5 are omitted in this post as these were over areas remote from the crash location.

On June 1, the Brazilian search grid was possibly/probably just to the right of the crash location, but the French clearly overflew the location. What are the possible explanations for this?

a.) the French were blind, or searched under poor observational conditions.
b.) any objects floating on the surface had already moved to the west, beyond the search grid, in the current and drift by the time the French searched.
c.) objects had not yet reached the surface when the French overflew the location.
d.) the French gave erroneous data to Metron as to the spatial area covered by their June 1 search.

The reasonable conclusion would be that if you overflew a possible crash location within 24 hours of the crash and observed nothing, then that is not where you will find the airplane.

The Metron analysis notes that the ship Douce France searched in the area of the LKP on June 1, but does not give the track of the ship.

All the current and drift reconstructions for June 1 and probably for the several days following, now appear to be badly flawed. But again these also presumed that evidence of the crash should have been spotted by the overflights on June 1 if that was within the area the plane crashed.

Here are links to the Metron analysis:
http://www.bea.aero/fr/enquetes/vol.af.447/metron.search.analysis.pdf

And the analysis done by the drift group:
http://archimer.ifremer.fr/doc/00027/13777/10915.pdf

slats11

6th Apr 2011, 11:49

Hi SaturnV. There are plenty of questions here, and we still do not have enough information to answer them.

I strongly believe that the debris found was always floating somewhere. It was simply not found. Either they overflew it and missed it, or they were searching the wrong area.

Debris can be very easy to miss in rough seas - it is difficult to see amongst all the white water. If you look at the image of the divers recovering the VS however, it is clear that the sea was very calm (at least at that time). So that does not appear to be the explanation.

There was certainly overcast, and this would have hampered the search.

With a lot of ocean to search, they probably flew at a relatively high level ie they decided to maximise the area covered at the expense of resolution. They searched an area of approximately 120 x 60 miles over the first few days. This is a very substantial search given it is right in the middle of the ocean, so they were clearly trying to cover plenty of area. This makes sense. They didn't know what had happened or where to look, but would almost certainly have still been hoping for survivors. Therefore they likely conducted an extensive low resolution search (looking for life rafts etc and knowing they needed to be found quickly). So I expect they flew high with fairly wide tracks so as to increase the "searched area", but also increase the risk of missing something within that area (especially with cloud).

I guess they were just unlucky. The total amount of debris recovered is not a lot, and is easy to miss from the air when you don't know where to look.

Does anyone have any idea where the location supposedly is. They have hinted close to LKP, but I wonder if this is disinformation.

Chris Scott

6th Apr 2011, 11:52

Hi HN39 (and PFR),

Don't have a picture, I'm afraid. You will recall that PJ2, you and I had a short discussion the day before yesterday, ending here:
http://www.pprune.org/tech-log/395105-af-447-search-resume-150.html#post6350925

Earlier, not being a structures man, I tried clumsily to explain that the MLG is mounted on the rear spar, near the apex of a triangle formed at the point of intersection of a diagonal spar with the rear spar. The triangle formed (sometimes called the Bermuda Triangle) has as its base, I think, the keel of the centre-section fuselage. The MLG assembly, as you said, retracts inboard so that the tyres nearly touch the keel as they rest in the main-gear bay.

Again, I stand to be corrected by an airframes expert, but the triangles surrounding each of the main-gear bays must be about the most robust structures of the airframe. They may have survived the impact, protecting the main gear legs if they were retracted.

The photo of the R/H MLG on page 5 of the "Présentation BEA":
http://www.bea.aero/fr/enquetes/vol.af.447/point.presse.4avril2011.pdf
does not show anything inboard of the extended leg, but that does not necessarily mean that the visible part of the rear spar has broken away from the "Bermuda Triangle": we will find out later.

If the MLG was in the retracted position on impact, the disturbance to the MLG up-lock and the ripping-off of the lightweight main doors could allow the leg to drop suddenly or gradually into the position shown; perhaps like a gravity-extension procedure.

Chris

takata

6th Apr 2011, 12:04

Given the approx location of the debris, is the fact that the 02:10:30 (assume this is exact time of position acquisition, transmission 4s later) location was "slighty west 3NM" of track INTOL - TASIL now more significant.

Yes was discussed at length previously, and can only be answered for sure by the black boxes - but was the deviation deliberate or indicative of an earlier upset?
More likely, this small deviation from the flight plan was deliberate, certainly due to weather avoidance.

People should use the term "upset" more carefully. It means a loss of control when an aircraft flight attitude become dangerous (in pitch, roll...), something that can lead to a crash if not recovered immediately, and the crash may happen in a very short time following an upset (most of the time, 1 min or less). I don't think there is any precedence of an "upset" lasting 5 minutes (or more) before impact.

A deliberate change of the flight path (deviation, altitude, etc.) doesn't automatically qualify for an "upset" if the aircraft has not reached an "abnormal attitude". In this case, the attitude at impact doesn't tell much about an early "upset"... so far.

- if an engine flameout happened, was an ACARS to be transmitted?
- is a spinning "posture" making a stop to the satellite transmission (so, no more ACARS)?
- in the event of a vertical stabilizer rupture, is the cabin depressurization mandatory (and followed by an ACARS)?
A single engine flameout would be transmitted by ACARS, a dual would turn off the SATNAV transmission. I don't know if a spin may stop the transmission of ACARS, possibly not.
ACARS are transmitted when the toilets are clogged while there is plenty of systems linked to the Vertical Stabilizer (hydraulics, etc.). So yes, the maintenance computer would certainly notice the lack of it. The cabin pressure would be compromized or not, it depends of the airframe damages suffered in this case.

SaturnV

6th Apr 2011, 12:17

slats11, in the dark blue area 10-15 km WNW/NW of the LNP.

http://www.flightstory.net/wp-content/uploads/AF447-location-map.jpg

HazelNuts39

6th Apr 2011, 12:41

If 2 ADRs tell the same lie, the flight law will not degrade to ALT but remain to NORMAL.

Thanks for your reply, CONFiture. Sorry if my question was ill-considered. However, in minute 2:10 speeds became inconsistent, and FCPCs reconfigured to ALT 2 law, therefore low speed stability was operative.

In twelve out of thirteen cases of erroneous airspeed discussed in BEA Interim Report No.2, the airplane switched to alternate law. In one case, this switch was temporary. Nine cases of triggering of the stall warning were observed.

Apparently "low speed stability" was never activated in these events, and I would like to understand in what conditions it would be.

Regards,
HN39

BJ-ENG

6th Apr 2011, 13:15

These might aid visualisation:

Main Landing Gear, Airbus A330/A340 | Flickr - Photo Sharing! (http://www.flickr.com/photos/nostri-imago/2971977894/)

3D max airbus a330 landing (http://www.turbosquid.com/FullPreview/Index.cfm/ID/336827)

HazelNuts39

6th Apr 2011, 13:56

BJ-ENG;

Thanks for those helpful pictures.

Regards,
HN39

Bobman84

6th Apr 2011, 14:10

Sorry, but it is not what Jean-Paul Troadec said. This is a bad translation of "All the wreckage brought to the surface will be sent to France", which is quite different: they will only brought to the surface what part is needed in order to fully understand the crash; they won't need the full aircraft if the recorders are recovered and are telling enough about the cause.

Thanks for the correction. However if the recorders provide no answers (ie Swiss Air 111, TWA 800), then I would imagine most (if not all) of the wreckage would be raised as has happened before.

mickjoebill

6th Apr 2011, 14:16

Hopefully that bit of oarnge will not be too heavily silted or otherwise covered if it is lying free.
Then a good photo interpreter will have a fighting chance of picking it out.

Part of my emergency kit is a rescue streamer, comprising a 6 inch wide x 20 meter long ribbon of plastic.

I've suggested before that the FDRs should deploy material that can aid its recovery underwater or if buried in mud. The box could have an outer frangible shell containing very long ribbons of metallic material, or perhaps slow (50 day) release dye.

We seem to be able to track radiation to one part per trillion, perhaps there is a harmless isotope we can employ?

Mickjoebill

fantom

6th Apr 2011, 14:20

There are some pretty good images of the gear here:

3D max airbus a330 landing (http://www.turbosquid.com/FullPreview/Index.cfm/ID/336827)

PJ2

6th Apr 2011, 14:31

Chris, HN39;
If the MLG was in the retracted position on impact, the disturbance to the MLG up-lock and the ripping-off of the lightweight main doors could allow the leg to drop suddenly or gradually into the position shown; perhaps like a gravity-extension procedure.
I think both HN39 and you are reading the photograph of the right gear/wing/flap track correctly. That is indeed a flap track on the right side of the photograph, (confirmed). The diagrams below may help sort out the original question, Chris.

I am aware that another poster had considered the vertical water impact on the gear doors may have molded them to the high points of the retracted gear and, given the surface area of the MLG itself, (let alone the doors), presented to the water, that the gear would not have been forced to the down-position. I think that is a reasonable expectation, for the moment of impact.

One thing is for sure in this theory regarding why the gear is resting in the extended position: Against impact forces alone, the uplocks would not have held the MLG up, nor would have the MLG doors. They would have broken at impact.

It remains to be understood whether the impact with the surface of the water would have kept the gear up, at least until the wreckage began settling or whether the gear broke through such forces and extended at the moment of impact. I don't think it matters.

Alternatively, the crew threw the gear out as a last-minute attempt to regain control and the recorders will tell us that.

What is interesting is, it's extended and there are two reasonable notions why - impact forces or crew action. We will learn soon enough I think.

PJ2

http://i277.photobucket.com/albums/kk76/batcave777/MLGinstallation2011-04-06_064138-1.jpg

http://i277.photobucket.com/albums/kk76/batcave777/MLGandDoors2011-04-06_064221-1.jpg

takata

6th Apr 2011, 15:38

takata,
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.
Not 15 nm: maybe less than 5 nm from LKP.
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.

HazelNuts39

6th Apr 2011, 16:07

Perhaps we should look at the flap track too. BEA's Interim Report No.2 has pictures of it in extended and retracted position (pp.20-21).

Regards,
HN39

lomapaseo

6th Apr 2011, 16:57

For some more consideration in water impact on the landing gear door, take a look into the TWA800 photos. I seem to recall that there was quite a bit of discussion about one of the gear doors being mangled enough (later concluded from water impact) to feed the conspiracy folks

BOAC

6th Apr 2011, 17:19

What is the expert opinion on the likelihood of the gear remaining attached, if extended, in a ditching?

RatherBeFlying

6th Apr 2011, 17:28

PJ2

6th Apr 2011, 17:30

HN39;
Perhaps we should look at the flap track too.

Things to note, (as per BEA conclusion that the flaps were retracted):

The flap actuator/drive has extended the flaps in this photograph from the BEA Report No.2, (rotated and desaturated to make comparison easier). The lower photograph shows the same structure, (cropped from the larger one with the gear and rear-spar showing) and not only confirms that this is the right wing but shows the flap actuator in the "up" position, as per the drawing:

http://i277.photobucket.com/albums/kk76/batcave777/A330FlapTrackrotatedbw011-04-06_093240-1.jpg

http://i277.photobucket.com/albums/kk76/batcave777/Flaptrackdrawingflapsup2011-04-06_102229.jpg

http://i277.photobucket.com/albums/kk76/batcave777/rightlanding-gear-wing-flaptrackfinisheddetail.jpg

takata

6th Apr 2011, 17:56

If the MLG was retracted at impact, I woulod suggest the wing spar holding the gear leg swivel broke away from the structure fuselage taking the gear with it. Perhaps the wing could have pivoted up at impact and once headed down the gear swung to the down position. This is getting complicated.

Occam's razor leans to the crew dropping the gear in an attempt to regain control.

One may cite also H. L. Mencken: "Explanations exist; they have existed for all time; there is always a well-known solution — neat, plausible, and wrong."

In fact, without a modelization of the impact at sea, including a model showing the sinking of the remaining MLG parts down to 4,000 meters deep, with all surfaces implied, weight, buyancy and pressure effects, it would be very difficult to refer to any simple "law".

Maybe the MLG extended during its fall to the sea bed after being released by impact forces. So far, I'm not sure the MLG would be so preserved (especially the tires) if it had sustained the full initial shock without being stored.

mm43

6th Apr 2011, 18:31

Originally posted by takata ...
Maybe the MLG extended during its fall to the sea bed after being released by impact forces. So far, I'm not sure the MLG would be so preserved (especially the tires) if it had sustained the full initial shock without being stored.I would tend to agree with this assumption.

Two days ago, by PM to another poster on the same matter, I answered as follows:-

I think we need to consider the relative acceleration (deceleration) forces acting on various areas at impact, and what intact mass was behind them. By this, I mean we should consider the main spar (including the center tank) and then look at the mass and relative size of the MLG. On a mass/area basis the MLG will decelerate at a lesser rate than the main spar/wing area, and I think that is the basis for it being found in what could be considered the locked down position.

The DFDR will confirm this one way or the other.

HazelNuts39

6th Apr 2011, 18:34

PJ2;

I thought that in the Report pictures we are looking towards the trailing edge, i.e. wing on the left, flap on the right. (See pylon trailing edge at the far left of second picture). In the sea-bottom picture we are looking towards the leading edge, and I was searching at the forward end of the heavy track or beam, for the arm that is flush with the wing under surface in the first report picture (flap extended), and perpendicular to it in the retracted position shown in the second report picture. The drawing you show upsets my sense of orientation. Have I got it all wrong?

Regards,
HN39

sensor_validation

6th Apr 2011, 19:12

More likely, this small deviation from the flight plan was deliberate, certainly due to weather avoidance.

People should use the term "upset" more carefully. It means a loss of control when an aircraft flight attitude become dangerous (in pitch, roll...), something that can lead to a crash if not recovered immediately, and the crash may happen in a very short time following an upset (most of the time, 1 min or less). I don't think there is any precedence of an "upset" lasting 5 minutes (or more) before impact.

A deliberate change of the flight path (deviation, altitude, etc.) doesn't automatically qualify for an "upset" if the aircraft has not reached an "abnormal attitude". In this case, the attitude at impact doesn't tell much about an early "upset"... so far.

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?

PJ2

6th Apr 2011, 19:28

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

BJ-ENG

6th Apr 2011, 19:41

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

Airbus A330 Landing gear | Flickr - Photo Sharing! (http://www.flickr.com/photos/babak1/3628049574/)

BOAC

6th Apr 2011, 19:48

About 10^-9 - that's one firm vote for a clean a/c.

HazelNuts39

6th Apr 2011, 20:04

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

JD-EE

6th Apr 2011, 20:27

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!)

JD-EE

6th Apr 2011, 20:38

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.

milsabords

6th Apr 2011, 21:25

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

ChristiaanJ

6th Apr 2011, 21:49

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.....

PFR

6th Apr 2011, 22:01

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:

HazelNuts39

6th Apr 2011, 22:08

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.

Alison747

6th Apr 2011, 22:30

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?

PJ2

7th Apr 2011, 00:03

HN39;
what I mistook for engine pylon is the wing leading edge. ...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;
Any sense?
Completely.

KTVaughan

7th Apr 2011, 00:19

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......

Chris Scott

7th Apr 2011, 00:46

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...)

auv-ee

7th Apr 2011, 01:59

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......

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 (http://www.bea.aero/en/enquetes/flight.af.447/trajectoires/trajectoires.html)

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

mm43

7th Apr 2011, 02:48

Originally posted by Chris Scott ...
... 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.There are some indicators that will help, e.g.

Those bodies recovered were probably not strapped in at the onset of the upset.
Some abnormal 'g' forces were experienced throughout the upset period.
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.
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.

Machinbird

7th Apr 2011, 03:22

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.:}

CONF iture

7th Apr 2011, 03:49

Apparently "low speed stability" was never activated in these events, and I would like to understand in what conditions it would be
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.

Let the people who think 911 was an inside jobCall them any name you like but not before you study building 7 ...

Things to note, (as per BEA conclusion that the flaps were retracted
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.

JD-EE

7th Apr 2011, 03:59

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.

Machinbird

7th Apr 2011, 04:42

for much of this crash what is often called "meatware" was in control.Or perhaps attempting to regain control after the flying machine did something "off the wall" while iron mike was in control.
I wouldn't be surprised if the accident was found to be beyond the aircrew's control and experience.
(Other than by not flying into the weather in the first place, but we just do not know what they saw up ahead in their radar at this point.)

ZeeDoktor

7th Apr 2011, 05:20

There's a P in PIO... the meat as some of you call it. It took a pilot to rapidly deflect alternating full rudder until it snapped off, in the case of AA587, that was no autopilot causing it.

In fact, I don't think there has ever been an incident where an autoflight/autopilot system was ultimately responsible for the crash of an airplane, so if this system caused departure into an unsalvageable state, that would be a first.

Except, it's not a first for pilots to fly into the top end of a (major!) CB and losing it.

So, when you hear a stampede, think cows, not zebras (for those of us living where there predominantly are cows)...

Machinbird

7th Apr 2011, 07:08

There's a P in PIO...

Yes, and autoPILOT has a 'Pilot' in it.
As I stated earlier, it relates to PIO (but is not strictly PIO since no human hand is on the control.)

The human pilot uses variable gain strategies that can be unexpected or inconsistent, but an autopilot must follow its program.

A fly by wire aircraft is being flown by the autopilot essentially until it degrades to a flight law that is strictly pilot directed. Direct law on Airbus is such a pilot directed mode if I recall the system correctly.

Some of you may also be surprised to learn that the A320 aircraft has encountered lateral PIO on a line flight.

PIO is essentially a cliff like change in aircraft behavior that can be brought on by a number of factors. but incorrect gain in the control loop or excessive time lags at the effectors (control surfaces to you) often set it off.

For those that are curious about the subject, find a copy of "Aviation Safety and Pilot Control" published by National Academy Press, ISBN 0-309-05688-8. It is probably more understandable than the textbooks and is moderately priced.

BOAC

7th Apr 2011, 07:51

To be frank, I can't imagine why anyone would even think of putting the gear down if ditching seemed inevitable - Clive - your profile does not say much, but I guess you are looking at this from an engineering pov?

As others have said, it is more than possible that there was no 'planned' ditching. From a piloting pov, gear is sometimes lowered to try to control speed, and we do not yet know if this happened. It is useful to me to see if we can establish, from what we can see so far, the likelihood of the gear breaking free after impact rather than being extended AT impact. As with others, I believe the intial impact would have firmly 'trapped' the gear in the up position - whether inertia would have been sufficient to force it through the distorted doors and the uplocks is for physicists/engineers to say. I would hope that someone can see from the pics whether various jacks/angles etc indicate a simple 'free-fall' extension in the water or a planned one? Have we seen both main gear (and nose) yet?

Thank you for the information on gear stressing.

PJ2

7th Apr 2011, 07:59

CONF iture;

Originally Posted by PJ2
Things to note, (as per BEA conclusion that the flaps were retracted
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.
To answer your questions directly, the slats are not mentioned in either preliminary report, and it's not possible to determine the position of the slats from the photographs available on the BEA site. The wreckage and/or the DFDR will tell us soon enough but there is one possibility although it may be looking too deeply.

If the CAS is higher than 100kts, only the slats are available with a selection of Config 1. So for normal operation in flight, the position of the flap drive does not tell us anything about the slats.

If the CAS is below 100kts, selecting Config 1 will bring out the slats and also bring out the flaps to "1", (8.5deg), which is a normal takeoff configuration in, (Config 2 and 3 can also be used for takeoff).

Since the SFCC [for others not familiar, Slat - Flap Control Computer] gets its speed information from the three ADIRUs, [Air Data and Inertial Reference Units] and we know that there was a loss of airspeed data in the three ADRs we don't know how the SFCC would respond to a crew selection of flap 1.

When recovered, I think it will also be interesting to see what the THS [Trim Horizontal Stabilizer] or screwjack position is.

BOAC, we have seen both main gears but not the nosewheel.

PJ2

FlexibleResponse

7th Apr 2011, 08:16

Sadly, when it's all said and done, I think we might discover that the THS was at or close to the full NU position and in the "Use Manual Trim Only" mode at impact. This THS flight control degraded mode freezes the THS (trim wheel control only) and thus overpowers the pitching moment of the elevator control power via the sidestick for pitch changes.

I would suggest that at CG positions of say 34-38% which are typical of this type in cruise, that if the THS attained a position of full NU (or close to full NU), that it would cause the a/c to stall and furthermore not allow any chance of stall recovery...unless manual use of the pitch trim is used to reposition the THS in the ND sense.

Just speculation and my humble opinion of course...

hetfield

7th Apr 2011, 08:19

Since the SFCC [for others not familiar, Slat - Flap Control Computer] gets its speed information from the three ADIRUs, [Air Data and Inertial Reference Units] and we know that there was a loss of airspeed data in the three ADRs we don't know how the SFCC would respond to a crew selection of flap 1.

Hi,

isn't there a gear safety valve on A330 (I'm not current on 330)?
If so, where does it get the speed from?
What happens if no valid data signal?

Thx

HazelNuts39

7th Apr 2011, 09:54

The gear support structure has a shear pin designed to allow the gear to break cleanly in the event of a crash (9g longitudinal). This is so that it doesn't tear a great hole in the fuel tanks. The rear attachment of the pylon of wing-mounted engines incorporates a shear pin. It is designed to break when the engine is hit from below in a crash, and allows the pylon and engine to pitch up and depart over the wing, so as to avoid rupturing the fuel tank. I have difficulty imagining a similar function of a shear pin in the main gear support structure. Could you explain that in more detail?

HazelNuts39

7th Apr 2011, 10:28

if the THS attained a position of full NU (or close to full NU), that it would cause the a/c to stall and furthermore not allow any chance of stall recovery...unless manual use of the pitch trim is used to reposition the THS in the ND sense.
I believe you are referring to the A320 accident at Perpignan. In that accident the condition you describe developed because the AoA vanes were inoperative. If those vanes are free to move, automatic trim stops short of the stall AoA. I believe that at that point it is still possible to recover by normal use of the elevator.

Chris Scott

7th Apr 2011, 10:50

Quote from BOAC:
As with others, I believe the intial impact would have firmly 'trapped' the gear in the up position - whether inertia would have been sufficient to force it through the distorted doors and the uplocks is for physicists/engineers to say.

Just a reminder that each A330 MLG is a massive (heavy) structure - one of the biggest in the business - whereas the main door you are considering is suitably large, but made of a composite material about as light as fibreglass. I'm also not convinced that this material could wrap itself around the bogie (wheel truck) without shattering.

It is true that pilots occasionally use the gear as an airbrake, and - in extremis - even to provide lateral stability. A double-engine failure would probably prevent the gear being subsequently retracted for a ditching, due to the effect on hydraulics, even if the APU could be started in good time. Once down, therefore, it would probably have remained down. So: was it?

CliveL's 9g longitudinal shear-pin - or something equivalent - must prevent an extended MLG from compromising the rear spar (or my "Bermuda Triangle"). The fact that the rear spar and MLG are still visible ensemble in the BEA photo convinces me that the gear was not in the extended position at the end of the flight.

Chris Scott

7th Apr 2011, 11:15

Quote from Machinbird:
Some of you may also be surprised to learn that the A320 aircraft has encountered lateral PIO on a line flight.

Not uncommon, in my line-checking experience, due to what I have referred to (on another thread) as "sidestick abuse".

HazelNuts39

7th Apr 2011, 11:24

whether inertia would have been sufficient to force it through the distorted doors and the uplocks is for physicists/engineers to say.That it would do that is not in dispute, nor that the extended gear would separate at high enough forward speed. (Some "deep stall" or "flat spin" adepts postulate near-zero forward speed). What I find difficult to believe is that, if the gear was separated from the wing in more or less retracted position, taking large parts of its support structure with it, that it would end up in fully extended position relative to its support structure. Wouldn't the longitudinal 9g forces referred to by CliveL hinder its 'gravity extension'?

Regards,
HN39

grity

7th Apr 2011, 11:30

http://s7.directupload.net/images/110407/temp/dm8hdjdd.jpg (http://s7.directupload.net/file/d/2487/dm8hdjdd_jpg.htm)
if this strukture was part of the fuselage, then we mayby have a look of the middle seats, row 4 to 7
grity

OleOle

7th Apr 2011, 12:22

I guess in case of a retracted gear, it would depend on the A/Cs attitude during impact if upward or downward forces on the gear prevailed: the inertia forcing the gear down or the water pushing the main door an the gear up. The main door of the MLG would be one of the first parts of the A/C to hit the water.

BOAC

7th Apr 2011, 12:23

HN - I do not see it so clearly - the water, which is providing the retardation force is 'retarding' the gear also and may balance the 'inertia' of that. Only if you slowed the rest of the a/c leaving the gear unaffected by entry into the water would what you say be a 'definite'. I am sure the only certain way to know is to look at the various bits of the assembly in terms of what goes where on a 'normal' extension and what has sheared from what in line with CliveL's post. We could well just be looking a slow 'gravity' drop from a broken airframe as it made its way down 3900m.

ttcse

7th Apr 2011, 12:29

Regarding lack of occupant preparation...

mm43...

There are some indicators that will help, e.g.
Those bodies recovered were probably not strapped in at the onset of the upset.
Some abnormal 'g' forces were experienced throughout the upset period.
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.
Any ensuing spin/rotation would have introduced forces that prevented cabin crew and pax returning to their seats.
If it were primarily a deep stall with not much rotation, I would expect the 'g' forces to actually be fractionally less than 1g for most occupants.

I tend to think there was 'total suprise and shock' in the cockpit of loss of control (on top of loss of everything else that had already happened). Furthermore there was some understanding that the situation was priority#1, absolutely all effort must be applied to deal with it as quickly as possible, everything and anything else was totally moot until it was resolved.

tt

dvv

7th Apr 2011, 12:40

CONF iture, what about building 7? The NIST team has studied WTC7 quite thoroughly, and the study only confirms that the 9/11 conspiracy nuts are just that — conspiracy nuts.

[end of off-topic]

dvv

7th Apr 2011, 12:43

ttcse, the data from Pulkovo 612 confirm your expectations — the RPM were too low to produce any noticeable Gs.

Chris Scott

7th Apr 2011, 12:55

Quote from BOAC:
HN - I do not see it so clearly - the water, which is providing the retardation force is 'retarding' the gear also and may balance the 'inertia' of that. Only if you slowed the rest of the a/c leaving the gear unaffected by entry into the water would what you say be a 'definite'.

I wonder if you may be missing the point, BOAC, re the inertia of the heavy MLG; in particular the bogie/wheel-truck part (furthest from the pivot point). At the extreme VS thought likely, the deceleration of the belly, or centre-section, as it struck the water would be great, due to its large surface area and relatively low mass. The bogie has a high mass and small surface area, relatively speaking. The door underneath it is light and flimsy.

Chris

ZeeDoktor

7th Apr 2011, 13:18

Machinbird,

I happen to be intimately familiar with a number of PIO incidents on A3XX aircraft, and there are two factors that have been most important in these:

1. The A3XX sidestick (control) philosophy is not rate but attitude based, i.e. you move the sidestick to the side (and with it a caret on the PFD), and the bus will keep the attitude you demand by pointing that caret. So, for example, you move the caret in the right hand side of the PFD and the bus will maintain an attitude to follow that caret. Unlike an analog system, where you bank the airplane and it will maintain most of the bank on its own while you just keep the bank angle with what's essentially neutral controls.

2. When in direct law (i.e. the proverbial has hit the fan and you're one failure short of MAN PITCH TRIM ONLY), it'll start behaving like an airplane you learnt to fly on... and you're in effect flying a rate based control system, where your inputs have to be given differently.

While I have great respect for the beautiful autoflight systems airbus has developed, this very basic control philosophy is flawed in my opinion and could have been part of what unfolded in the middle of the night over the atlantic, inside a CB. Don't go there. We teach that at basic levels...

auv-ee

7th Apr 2011, 13:39

A BBC update from this morning claims that there will be a meeting with families in "two weeks".

BBC News - Conflicting emotions of Air France crash relatives (http://www.bbc.co.uk/news/world-europe-12989002)

wes_wall

7th Apr 2011, 14:15

Are any further photos expected. Lack of additional information regarding the site or discovery is loudly silent. A sign of things to come?

Chris Scott

7th Apr 2011, 14:56

Quote from ZeeDoktor:
The A3XX sidestick (control) philosophy is not rate but attitude based, i.e. you move the sidestick to the side (and with it a caret on the PFD), and the bus will keep the attitude you demand by pointing that caret. So, for example, you move the caret in the right hand side of the PFD and the bus will maintain an attitude to follow that caret.

I'm afraid that is all incorrect!

If you move the sidestick to the side in the air, you are requesting a roll-rate proportional to the displacement. If you then centralise the stick, the bank will be maintained.

This does not apply, however, on the ground. On take-off, it is stick-to-surfaces until about one second after lift-off, when the white cross (that you call a 'caret') disappears from the PFD for the duration of the flight.

(Pitch control is a bit more complicated to explain, but intuitive.)

In the air, PIO is the result of the pilot impatiently or nervously changing the stick position before his/her previous selection (stick-position) has had the time to be effected. In that respect only, it's not so different from conventional aeroplanes. The trick is to use thumb-tip and index-fingertip for roll control; thumb-tip and middle-fingertip for pitch.

Chris

PS (by Edit)
You are right to say that, in Direct Law, it's a bit like "an airplane you learnt to fly on", i.e, roughly stick-to-surface. That does not mean "a rate-based control system", however. You are right that it feels very different from the higher laws, and is rarely practised except for short periods in the simulator.

takata

7th Apr 2011, 15:01

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?
0200 and 0210 are the only two positions transmitted during this time window, it doesn't tell much about the flight.
Who know, today, her route and what exactly F-GZCP radar pictured on her path?

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?
The real question is how and when she departed from altitude, not that she actually departed.
My point is that other factors can't be ruled out at this point as there is absolutely no certaincy about the time and nature of the "upset" causing the crash. Many ACARS were sent from 0210 and it is very likely the start of a following sequence leading to one or more catastrophic events... but it doesn't mean an immediate unrecoverable upset!

See below about an aicraft falling from the sky like a "leaf" in 5 minutes (or more). One can make his own opinion.

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?
No-one can so far explain why switching at 0210 to ALT2, due to unreliable airspeed, would cause an immediate upset as none of the similar events ever caused one! After being switched to manual control, any manual imput could cause one in relation with other unknown factors (workload, weather, technical, etc.)

But surely the assumed proximity of the LKP to crash site suggests one major unrecoverable incident?
Yes: one upset... or more than one!
In my opinion, the attitude at impact and the rate of descent is not ruling out more than one incident!

9 unrecovered upsets from cruise level (data Metron Inc., BEA study):
Date.........ft/mn..Level..Dur....Dist..Type....-> Cause
07.12.1995..32,000..FL310...57s...8 NM..TU-154B -> Spiral / roll upset
19.12.1997..29,000..FL350...75s...5 NM..B-737.. -> Unknown
21.12.2002..27,000..FL180...40s...2 NM..ATR 72. -> Stall / icing
19.11.2001..26,000..FL260...59s...4 NM..IL-18V. -> Spiral / diving
01.01.2007..20,000..FL350..105s...9 NM..B-737.. -> Roll upset
15.07.2009..16,000..FL240...90s...5 NM..TU-154M -> Spiral / roll upset
22.08.2006..14,000..FL390..166s...3 NM..TU-154M -> Stall-spin
23.03.1994..12,000..FL310..156s...3 NM..A-310.. -> Spiral-spin / roll upset
16.08.2005..12,000..FL310..210s..17 NM..MD-82.. -> Stall

AF447 - timed cases for one upset from cruise level:
01.06.2009..35,000..FL350...60s..~5 NM..A-330 (1 mn)
01.06.2009..23,000..FL350...90s..~5 NM..A-330 (1.5 mn)
01.06.2009..17,500..FL350..120s..~5 NM..A-330 (2 mn)
01.06.2009..12,000..FL350..180s..~5 NM..A-330 (3 mn)
01.06.2009...9,000..FL350..240s..~5 NM..A-330 (4 mn)
01.06.2009...7,000..FL350..300s..~5 NM..A-330 (5 mn)
01.06.2009...6,000..FL350..360s..~5 NM..A-330 (6 mn)

Recovered upset: B747 China Airlines (17.02.1985)
1. 1014:50 -> 40,442 ft -> roll upset (4.8 G pull up)
2. 1015:23 -> 30,132 ft -> unreliable attitude data (5.1G pull up)
3. 1017:15 -> 9,577 ft -> recovery & climb

Total upset duration: 145 sec.
1..10,310 ft lost...33 sec..18,700 ft/mn
2..21,155 ft lost..112 sec..11,300 ft/mn
3..31,465 ft lost..145 sec..13,000 ft/mn (average)

"Upset" definition (from “2008 - Airplane Upset Recovery Training Aid Revision 2”):
An airplane upset is defined as an airplane in flight unintentionally exceeding the parameters normally experienced in line operations or training. In other words, the airplane is not doing what it was commanded to do and is approaching unsafe parameters.

While specific values may vary among airplane models, the following unintentional conditions generally describe an airplane upset:
• Pitch attitude greater than 25 deg, nose up.
• Pitch attitude greater than 10 deg, nose down.
• Bank angle greater than 45 deg.
• Within the above parameters, but flying at airspeeds inappropriate for the conditions.

bearfoil

7th Apr 2011, 15:05

Given the 'lenient' limits granted the Autopilot in cruise, and bunk weather, it is reasonable to assume the pilots were established in "acceptable" autoflight until sometime just before the stream of ACARS signalled what is likely an already upset aircraft. The mechanical (A/P) limit having been breached, the pilots were left with a rapidly manouvering airframe with which they had not had time to acommodate their 'meat-think' (:ugh:). As per ZeeDoktor, the transition between a docile machine and one that requires a stable set of anticipatory thinking only, the conclusion is quite apparent. (That this is a flaw in AB philosophy)

If impact was 6.5nm away from last known position, it would need a bit more forward travel past LKP to avoid considering the a/c dropped like a stone at an average AOD of 45 degrees. This means a turn (back) of some description, and most likely pilot input. This excuses to some extent the searches assumption of distance away from LKP.

A "flat impact" is concluded, but this is not to say that she wasn't in some Pitch extremis just prior to impact. If a tail-first hit, the VS driving downward in its aft spar axis explains the lower corner (Rudder) damage, and could allow for the reasoning that concluded the VS landed away from the crash site.

This also allows for some 'time' for the Mains to sever their stowed position downward, just prior to contacting the sea essentially vertically, and perhaps without a horizontal vector to speak of.

takata

7th Apr 2011, 15:34

... As per ZeeDoktor, the transition between a docile machine and one that requires a stable set of anticipatory thinking only, the conclusion is quite apparent. (That this is a flaw in AB philosophy)

Well, Bearfoil... maybe?
But, at this point, finding only such an (obvious) conclusion is much more telling about what you will find very relevant in support of your own theories rather than giving a single hint about this manufacturer philosophy.
Everything opposed to your pre-fabricated opinion should otherwise being considered as a "cover-up" as you already mentioned many times this very same obvious and single conclusion (I remember that you even mentioned that Airbus/BEA would not try to recover this aircraft as they already know pretty well what caused this loss).

Nothing very new.
:)

bearfoil

7th Apr 2011, 15:39

Do not put words down I have not said. At no time did I suggest the authority would not search, neither have I made a conclusion about this accident. I have a strong opinion, but am ready to accept any evidence that shows other than that. I resent your attempt to make a personal judgment outside the context.

takata

7th Apr 2011, 15:54

neither have I made a conclusion about this accident
Then, no offense. I'm very sorry: my very poor English is certainly not able to cope with your high syntax level that everybody can perfectly read in your post just above. But forgive me if I won't search into the hundreds of posts about this crash to find out what I mentioned myself. I certainly can be wrong or I just confused you with someone else who expressed, post after post, exactly the same view ("strong opinion") as you!
:}

Loose rivets

7th Apr 2011, 16:49

it would need a bit more forward travel past LKP to avoid considering the a/c dropped like a stone at an average AOD of 45 degrees. This means a turn (back) of some description, and most likely pilot input.

That just made me mindful of a time just north of the Pyrenees when I was trying to tell the guy on the ground that we had little idea of our position. We were on a 22min beacon-to-beacon leg, and 47 mins later we finally passed that beacon. For a lot of the time, our progress had a high vertical component - it was nothing short of chaotic for well over an hour, but there was never a time we turned intentionally off course.

If, as it seems is the case, this poor crew found themselves inside a vicious CB, only a second by second analysis of their exact position will throw much light on those terrible minutes. I'm not sure even the FDR will reveal that much detail as only inertial summing could reveal the details of the decent.

Is it hoped there will be that kind of detail in the FDRs?

auv-ee

7th Apr 2011, 16:56

This just posted at:

Sea Search Operations, phase 4 (http://www.bea.aero/en/enquetes/flight.af.447/sea.search.ops.phase.4.php)

Information, 7 April 2011
The team on board the Alucia will complete vehicle operations on Friday. The vessel will leave the search area on Saturday 9 April and should reach the port of Suape (Brazil) on the morning of Tuesday 12 April.

Chris Scott

7th Apr 2011, 17:04

Does this mean they have decided that all the wreckage has been found, I wonder, or that they have located all they need - including all three recorders?

wes_wall

7th Apr 2011, 17:09

The vessel will leave the search area on Saturday 9 April

Does this mean that the location of vital components have been identified and mapped, and therefore no further activity is required by them. Seems there is an abundent of information known but not released.

PJ2

7th Apr 2011, 17:36

The door underneath it is light and flimsy.
Not to put too fine a point on this aspect of the discussion...

Yes, agree with your comments. Further to your point in your response to BOAC, (who, I should acknowledge, has previously mentioned the notion that the gear may have free-fallen slowly, after impact, and during the slow descent to the bottom)...

The MLG doors are made from Carbon-Fiber Reinforced-Plastic (CFRP) with a NOMEX (http://www.corecomposites.com/nomex.cfm?family_id=3) honeycomb core. The belly-fairings are similarly constructed,(the cabin material is the same NOMEX honeycomb structure but with a fibreglass surfaces and aluminum extruded edges/fastenings.) So while light and strong, they are indeed "flimsy" in the sense that direct, high-impact forces would break the material as we have seen in the recovered wreckage. IOW, this material will not "bend and conform" to the shape of the gear and so impede a free-fall, but would likely shatter, and provide no resistance whatsoever to the massively-heavy MLG structure. That the uplocks would break at impact has already been discussed.

Again, when the MLG actually free-fell is immaterial - it did, and it, and its dressings have remained attached to the rear spar and support structure. I think it is more important to rule-in/rule-out the crew extending the gear. It is a debate until the recorders tell us.

I mentioned the THS (#3129 (http://www.pprune.org/tech-log/395105-af-447-search-resume-157.html#post6355507)) not in relation to the Perpignan accident (AoA problem) but because I don't recall if/where in the three major AF447 threads, the failures of ADIRUs 1, 2 & 3, which provide input to the FCPCs which control the THS signals, have been discussed.

The aircraft went from M0.82 (or turb penetration speed of M0.78) to about M0.60/198kts (roughly the stall speed under the circumstances). I've tried that in an A330 simulator and it takes long time just to lose 80 to 90 kts in level flight with engines set to IDLE, (not suggesting that engine thrust was idle for AF447, I'm just describing the "loss of speed" experiment). So something else intervened to advance the loss of speed and approach to stall. What, and why?

There is always a causal path to a loss of control whether it resides primarily in human factors, (the known ones such as rule-breaking, distraction, overwhelmed, competence etc), or resides in technical/system causes or, (more likely) a blend of the two. As many have pointed out over the length of these threads, these have always been the primary questions.

PJ2

ACLS65

7th Apr 2011, 18:06

@wes_wall (http://www.pprune.org/members/45869-wes_wall):

It sounds like they are just switching gears from localization to recovery, not that they are done yet.

From the same BEA site:

"The operation would thus be divided into two phases:

a localisation phase (phase 4) financed by the industry, which will use a relatively small and thus less expensive vessel;
a phase of detailed observation and recovery if the wreckage is discovered (phase 5), financed by the French State."

"Phase 4 will come to an end upon detection and identification of the wreckage by the BEA, and at the latest, at the end of the third stage in July. If the wreckage is found, the BEA will immediately launch phase 5, involving detailed observation and recovery, by calling on one of the three vessels pre-selected in the meantime, which will be chosen on the basis of their availability at the time of the discovery."

The New Undersea Search Campaign (http://www.bea.aero/en/enquetes/flight.af.447/new.undersea.campaign.php)

takata

7th Apr 2011, 18:44

Does this mean that the location of vital components have been identified and mapped, and therefore no further activity is required by them. Seems there is an abundent of information known but not released.
No surprise. The area is small, flat and sandy and they've got 3 UAVs available for the mapping... But no ROV.
They will need the ROV to go further and for looking inside the wreckage.

You are right about the "rumors" from those sources "close to the investigation" (Governement, BEA, Justice, experts, families, journalists...)
One sample (all unverified):
. the wreckage is very close to LKP (straight down);
. there is still large parts of the cabin well preserved;
. some corpses still attached to their seat;
. some corpses lying around the wreckage;
. the BEA is incompetent, having searched mostly at the wrong place...
:ugh:

Chris Scott

7th Apr 2011, 18:49

ACLS65,

The only news here is how early they will have been able to complete Phase 4, and one must assume that the rest of today (Thursday) and tomorrow will be spent completing routine photography and the inventory of items in the debris field (or fields) already discovered.

They must have located the DFDR and CVR, and presumably the QAR (although the latter is not designed or located to be crash-resistant).

But their announcements certainly are, err..., somewhat restrained.

mm43

7th Apr 2011, 19:53

Does this mean that the location of vital components have been identified and mapped, and therefore no further activity is required by them.They will have completed high resolution sidescan sonar flights over the area, then reverted to camera fly-by of the debris field, probably taking shots from all four quadrants. They are obviously confident that all major components of the aircraft have been accounted for and that they have sufficient information to allow the recovery operation (Phase 5) to proceed in due course.

This operation has certainly proved the operational capability of the REMUS 6000 AUVs provided by the Waitt Institute for Discovery (2) and the Geomar Oceanographic Institute, Germany (1), and operated by the Woods Hole Oceanographic Institution's REMUS team.

asc12

7th Apr 2011, 20:15

Can anyone enlighten us about the resolution of sidescan sonars? The BEA images show that the Remus AUV was using 120kHz. What sort of spatial resolution could one expect at this frequency? And how is the data displayed?

In other words, what are we really looking at in this image (http://www.bea.aero/fr/enquetes/vol.af.447/images/msn109a700m.jpg)? Are the redder parts colored to represent things that are physically located higher, or have a stronger return, or something else? What causes objects to have a stronger return with sonar? At 120kHz, how small an object could one expect to pick out?

[sorry-- this was posted in another thread but is more relevant here, I hope.]

dufc

7th Apr 2011, 20:57

Is it conceivable that the vessel would leave the area without having located the structures likely to contain the various recorders?

I would have thought that this was one of the main aims of the project and the return to port to me indicates either location of the recorders or a high certainty that they are within whatever sections of the fuselage have been found on the bottom of the ocean.

If this was not the case would the search not continue....?

deSitter

7th Apr 2011, 21:04

Thanks for this extremely interesting thread. Remember the first thing found floating some distance from the main site was the neatly severed fin. There was much talk about the merits of composite materials and AA587 in New York (an A300).

All the talk about stalling and loss of stability brings me back to one of the first points made - did the airplane lose its VS and then go out of control, did it lose it from aero-forces during a spin/dive, or from impact with the water? Could the fin have been knocked off by an encounter with a huge up/downdraft?

-drl

fantom

7th Apr 2011, 21:09

There seems to be a slight misconception here about lowering the gear.

You don't normally lower the gear to get the speed down; but you (might) lower the gear to increase the rate of descent in Emerg Descent. I emphasise 'might'.

At low altitude (arriving for landing, say) you might lower the gear earlier than planned if you are a bit high on speed - there was a reference earlier to using gear as airbrakes.

Quite different situations.

bearfoil

7th Apr 2011, 21:14

One might also consider lowering the gear to get the nose down.

BOAC

7th Apr 2011, 21:39

Let's not get into pages of why we put the gear down, eh? We are merely trying to solve CliveL's dilemma about 'lowering the gear if you are to ditch', and I was simply hoping someone could identify from the images whether there were signs of a deliberate extension from jack positions etc. It really is not that important. I think it is fairly evident that ditching might not have been planned from the apparent state of the pax/cabin preparation, and no, CliveL, one would not normally lower the gear in a ditching scenario unless, perhaps a greater over-riding need arose.

To quote PJ2
"Again, when the MLG actually free-fell is immaterial - it did, and it, and its dressings have remained attached to the rear spar and support structure. I think it is more important to rule-in/rule-out the crew extending the gear. It is a debate until the recorders tell us."

Unless someone can tell from the pictures, we will indeed have to wait.

SaturnV

7th Apr 2011, 21:47

dufc, it may depend on the capabilities of the retrieval ship, on whichever one the government selects. WHOI has several months remaining on its contracted period, so it is not as if they need to rush off to another task.

And WHOI may still be needed. When a Canadian Coast Guard cutter was pulling up a forward cargo door (considered a vital item) from Air India 182, the line snapped as it was about to be hoisted aboard, and the door sank back to the abyss, never to be found again.

I would expect that once retrieval starts, there will be a small flotilla of vessels on the scene, including the French navy, and possibly the Brazilian navy.

auv-ee

7th Apr 2011, 23:16

Can anyone enlighten us about the resolution of sidescan sonars? The BEA images show that the Remus AUV was using 120kHz. What sort of spatial resolution could one expect at this frequency? And how is the data displayed?

The speed of sound in water is approximately 1500 m/s (it varies in important ways with temperature, salinity and pressure, but that is not relevant to this question). The wavelength at 120kHz is thus: 1500m/s / 120kHz = 12.5mm. How many wavelengths across does an object need to be to detect it? That's hard to say, because it depends so much on the reflective and resonant properties of the target. I'm not an expert in this area (same math as radar, so maybe others will contribute), but I would say that 10-100 wavelengths would likely be detectable; so that is roughly 10cm to 1m. That's for detection: having a strong enough echo to detect above the noise and clutter..

Resolution is a different question. Along-track resolution is determined by the beam width of the transducer and the motion of the vehicle between pings. The second is easy: at 600m range, for conventional side-scan, the sound has to travel 1200m to leave the vehicle, reflect from a target and return. With some dead time between pings, that makes about 1 ping/second. In that time the vehicle has moved 2 m, so that will be the best possible along-track resolution (unless the sonar is chirped or uses some other way of getting more sound in the water). I don't know the beam width of the system they are using, but it is likely about 0.5degree; that gives about a 5m wide beam at 600m, so that is another limit. Remember that this is about resolution, not detectability.

The other resolution limit, not usually the controlling limit, is the across-track resolution; that is, in the direction between the target and the vehicle. This is limited by some transducer and wavelength considerations, and also by the number of samples, in time, made of the returning echos. Typically that is between 250 and 2000 (what the eye can see), and so this is easy to make better than the along-track resolution.

The data are commonly displayed as an image built up of scans where each scan is the echo intensity (normalized for range and absorption) as a function of range, or more accurately: time scaled and corrected to represent range. Each scan line is perpendicular to the motion of the vehicle (hence the term "side-scan", the sonar is looking to the port and starboard sides using two separate transducers and receivers).

In the released image, you can see that the vehicle proceeded in an "up" or "down" direction with respect to the image (nearly north or south, note the dark bands under the vehicle where there are no immediate reflections), and that the targets are smeared more in the along-track direction than in the across-track direction. I think that picture is a mosaic of views from both sides of the debris (there is more than one dark "nadir" under the vehicle).

In other words, what are we really looking at in this image (http://www.bea.aero/fr/enquetes/vol.af.447/images/msn109a700m.jpg)? Are the redder parts colored to represent things that are physically located higher, or have a stronger return, or something else? What causes objects to have a stronger return with sonar? At 120kHz, how small an object could one expect to pick out?The only thing the side-scan can measure is the time and amplitude (and phase, for some systems) of the returning echos. The colors chosen by the operator to represent the amplitude could be anything. In this case the darker colors represent weaker signals, and the brighter colors represent stronger signals.

Return strength from a target is a very complicated topic. It depends on the material, the internal cavities, the shape, the size the orientation toward the receiver, and many other factors. For this long range phase of the search, they were just looking for targets that did not look like the background everywhere else. While the resolution at 120kHz may not be great, higher frequencies are absorbed by the water and thus the maximum range gets shorter. A 55 gallon oil drum would be a reasonable target to detect using 120kHz at 600m range; it would not be possible to resolve the shape and thus identify this target at that frequency and range, using the side-scan presently installed on these vehicles.

After acquiring the debris field they will have surveyed it with higher ferquency side-scan, at intentionally shorter ranges (better along-track resolution). Of course, by now they will also have about 100,000 photos, most of which will show mud with no A/C parts, but the rest will be interesting.

Machaca

7th Apr 2011, 23:31

deSitter:
...the neatly severed fin. There was much talk about the merits of composite materials and AA587...Hardly a neat separation -- the forged lugs were ripped out from between their forged hoops!

http://i337.photobucket.com/albums/n385/motidog/Lugs01.jpg

A330 section 19:

http://i337.photobucket.com/albums/n385/motidog/Section19-01.jpg

Example from A400M:

http://i337.photobucket.com/albums/n385/motidog/Lugs3-1.jpg

The talk regarding composites and AA587 were/are irrelevant to this accident.

Chris Scott

7th Apr 2011, 23:48

auv-ee,

Just to say thanks for that really excellent piece above - I understood all of it (I think)!

takata,

Did I see you and mm43 contributing to that "other place"? Dommage!

PJ2

7th Apr 2011, 23:54

Machaca - absolutely superb photos, thank you, - they explain/demonstrate a great deal.

PJ2

ZeeDoktor

8th Apr 2011, 00:09

Chris,

why do you say I'm wrong when you're emphasising what I said?

While you initially give a rate based on stick displacement, the airplane then holds the commanded attitude (with centered controls!). It hence does NOT behave like an airplane we all learnt to fly on, also the only airplanes we ever had any *real* upset recovery training in.

I rest my case, one should never implement systems which behave differently from what one learnt initially. We had a good example of that in Zurich not too long ago when a Saab 340 stuck it in near vertically shortly after takeoff, the Moldavian pilot had learnt to fly on AH's where the airplane symbol moves rather than the horizon. So he corrected the wrong way for an overbank condition.

mm43

8th Apr 2011, 00:50

Chris Scott,

Did I see (takata) and mm43 contributing to that "other place"? Dommage!

It was a "shameful" event! But shame on you for looking!

Sometimes a bit of "fact" injected into a nonsense thread will bring it to heel, and I suspect takata will agree.

Lazerdog

8th Apr 2011, 00:55

Machaca... Do you have any photos of the static dissipator tufts on the rudder after the vertical stab was recovered? It appeared in at least one photo during its recovery that several of the tufts were missing. After doing lightning analysis at a major avionics company, I think a possible root cause that should not be ruled out is high altitude lightning. A likely attach point forward could have been near the upper TCAS antenna. If the stroke traveled to the inside of the fuselage, a host of failures could have occurred. I know this is a stretch, but in the NASA F-106 studies that were done decades ago, the energies involved are very large. Recovery of electronic system components will certainly be valuable in the analysis.

Graybeard

8th Apr 2011, 01:21

Along that line, Lazerdog, nobody every gave a good explanation for the TCAS Fail in the ACARS message. Everybody just dismissed it as due to the pitot fails.

Again, a pitot fail will not affect the TCAS. A static fail will cause the transponder to revert to essentially Mode A function only, which will cause the TCAS to shut off, not Fail. Three things come to mind for a TCAS Fail:
Processor Fail
Lower TCAS antenna
Upper TCAS antenna

GB

Machinbird

8th Apr 2011, 01:29

Takata
No-one can so far explain why switching at 0210 to ALT2, due to unreliable airspeed, would cause an immediate upset as none of the similar events ever caused one! After being switched to manual control, any manual imput could cause one in relation with other unknown factors (workload, weather, technical, etc.)Perhaps this is stated backwards.
"Why would an upset cause unreliable airspeed indications??", might be a better question.
The answer is in the link below and subsequent discussion.
http://www.pprune.org/6210302-post2647.html

Note: With the discovery of the wreckage virtually under its 2010 position, the case for an immediate loss of control becomes stronger. The key question becomes how can it happen?

mm43

8th Apr 2011, 02:19

Originally posted by Graybeard ...
... nobody ever gave a good explanation for the TCAS Fail in the ACARS message. Everybody just dismissed it as due to the pitot fails. I seem to remember that was dealt with by the BEA in Interim Report #2. From memory, the TCAS has an inbuilt function which checks that the static port pressure is valid and that the calculations that result fall within predetermined values. Should the TCAS detect erroneous values, it fails.

I need to check the report for accuracy, but the above is near to it.

JD-EE

8th Apr 2011, 02:30

Chris, BOAC, let's take a wider view rather than a view immediately local to the MLG.

The belly slaps into the ocean. That provides a sharp dowards jolt to the gear. Perhaps tens of milliseconds later the wing becomes involved near the wing roots. There is a slap on the MLG doors mashing them upwards and likely shattering them. So far nothing has touched the MLG itself except a sharp deceleration. Perhaps it has already broken loose from its uplock and started to descend. At some point the MLG doors, whole or partly shattered touch the MLG. The doors are not nearly strong enough to contain the momentum of the MLG. It simply mashes through the doors or forces them (whole or not) out of the way as momentum pivots it to an extended position. Note that about the time the MLG hits the water the large wing surface is going to slap on the water providing a very sharp deceleration to the wing structure. (I'd not be surprised if the spars folded
up in the middle.) The MLG tries to keep going. The pivot point can't. But the truck and tires end can continue so it all pivots open.

I'd be surprised if the MLG was still neatly stowed. It would indicate a severe (too heavy) over design on the uplocks. I would expect to see a lot of torn and bent materials in this area due to these relative forces.

Mechanical Engineering is not my primary field. But visualizing the sequence simply requires a picture of the plane (Wikipoodle has some) and an ability to think in very small steps. Look at it and try to pretend you're seeing, in your mind's eye, that you are watching an extreme slow motion cutaway picture of the events.

jcjeant

8th Apr 2011, 03:05

Hi,

When we have the exact location of the wreck .. it will be interesting to compare his position with the position of the vertical tail when discovered
Could there be from those data whether this vertical tail detached on impact or before?
Is the distance between the wreckage and the vertical tail is logic with reference to wind speeds and current in this area?

mm43

8th Apr 2011, 03:35

Originally posted by JD-EE ...
The belly slaps into the ocean. That provides a sharp dowards jolt to the gear .... I thought I had proposed a similar look at the relative forces involved in post #3106 (http://www.pprune.org/tech-log/395105-af-447-search-resume-156.html#post6354572), but may be it got overlooked, because I believe that Chris Scott came up with a similar scheme a little later in post #3143 (http://www.pprune.org/tech-log/395105-af-447-search-resume-158.html#post6356055).

JD-EE

8th Apr 2011, 03:49

Oh nuts. That's a line I meant to edit and I posted it too soon. There is no real downward jolt to the wheel. There is a massive upwards jolt to the fuselage which would transfer to the wings as an initial upwards jerk on the hinge assembly and the uplock assembly.

That aside, your comment should not have been glossed over. It is a key to the landing gear down issue. It needs to be looked at step by step with the forces applied to the MLG assembly's various attachment points at least noted if not analyzed.

(Sometimes ya gotta mention things here several times to get them noticed.)

mm43

8th Apr 2011, 03:54

JD-EE

I sometimes prefer to be invisible.:ok:

My vote is that the gear was never lowered, and flaps and slats were also not extended.

JD-EE

8th Apr 2011, 04:00

And that raises an interesting naif style off topic question.

Presume you have something like an A330 falling like a leaf in some form of level stall with minimal forward velocity. What happens if you could extend flaps on ONE side only. Could that be guided into a new configuration from which escape from the stall would be possible?

I think it would tend to toss the wing with no flap extension downwards. Then the tail would point the nose downwards at a bloody scary angle. But that might allow enough airspeed pickup to allow the plane to get engines started and pull out to level flight.

Graybeard

8th Apr 2011, 04:07

I seem to remember that was dealt with by the BEA in Interim Report #2. From memory, the TCAS has an inbuilt function which checks that the static port pressure is valid and that the calculations that result fall within predetermined values. Should the TCAS detect erroneous values, it fails.

Altitude and its valid is transmitted on a data bus from the Air Data part of the ADIRU to the transponder. The transponder transmits own altitude on another data bus to the TCAS, and over the air to ATC and other aircraft.

Airspeed is not part of the TCAS calculation. TCAS uses only altitude difference, distance, and rate of closure.

If altitude data is flagged as failed, it does not cause a failure of the transponder or the TCAS. The transponder reverts to Mode A, and the TCAS annunciates OFF.

A TCAS Fail message means the TCAS has failed; it does not mean that it is turned off by an input fail or pilot selection.

mm43

8th Apr 2011, 04:23

Originally posted by jcjeant ...
Is the distance between the wreckage and the vertical tail is logical with reference to wind speeds and current in this area?It could be possible provided half the assumed current velocity is used. However as pointed out in the past, no one knows or knew at the time what the current was really doing. I am sure that attempts will be made to clarify where and how the V/S and all the other recovered debris made it to where they were found.

The position of the V/S in relation to the bodies found at the same time was reasonable, i.e. the leeway made by the V/S was greater than that of the bodies.

A standardized oceanic search and rescue protocol needs to be established to ensure in future that an initial aerial search doesn't make the same mistakes made with the AF447 accident.

mm43

8th Apr 2011, 04:38

Originally posted by Graybeard ...
TCAS uses only altitude difference, distance, and rate of closure.Not disputing any of that, other than in the case of this TCAS, it had a self checking routine that performed a "credibility" test.

From BEA Report #2:
This message indicates that the TCAS is inoperative. Without an associated fault message, it could be the consequence of an electrical power supply problem or of an external failure. Amongst the possible external failures, only one is compatible with the CFR received. This is a monitoring process internal to the TCAS which applies to the standard altitude parameter. The latter is received from the active transponder (it can thus be the altitude elaborated from ADR 1 or 2) and is submitted to a “credibility” test. In actual fact the TCAS elaborates an altitude prediction that it compares permanently with the altitude received. When these two parameters move too far apart, it stops operating and generates this ECAM message. Once the altitude becomes “credible” again, normal operation resumes and the message disappears.We don't know if normal operation ever resumed. I doubt it.

vanHorck

8th Apr 2011, 05:49

In post 3120 ( http://www.pprune.org/6355188-post3120.html )there was a detail of some weather overlaid the LKP and a question was asked to Tim Vasques about his interpretation of these colours.

Tim is too busy to follow the thread but his statement about the colours is as follows:

I haven't had time to research the incident since 2009, so I don't have anything to add. I saw your link and the question about colors. The colors shown indicate overall atmospheric radiance in the 10.7 micron infrared window. This would be a measure of cloud top temperature, with the reddest colors the coldest. This has a closer relation to updraft location than it does precipitation intensity, so I would be more prone to estimate turbulence from it than rain, though they will likely be closely associated in the tropics where the atmosphere is weakly sheared.

Tim

Tim produced a magnificent report about the weather conditions which were in place at the time of the accident

HazelNuts39

8th Apr 2011, 06:44

BEA Interim Report No.2, 1.6.11.6 Consequences of a drop in the measured total pressure, explains that a sudden drop in measured total pressure entails a sudden drop (300 ft) in indicated altitude, due to the correction of the static pressure.

It would be of interest to know the tolerance of the “credibility” test of the monitoring process internal to the TCAS which applies to the standard altitude parameter. It shouldn't be too difficult for the BEA to get that from the TCAS manufacturer.

Regards,
HN39

HazelNuts39

8th Apr 2011, 08:46

For the interested reader there is more information in BEA Report No.1 Appendix 1 and Report No.2 Appendix 3. The first reference has this Note:
See the MétéoFrance website (http://comprendre.meteofrance.com/pedagogique/dossiers/science_et_techniques/the_satellites_meteorologiques?page_id=2802)
for details on the analysis of satellite images.

Chris Scott

8th Apr 2011, 12:35

Hello ZeeDoktor (re posts #3174 and #3144),

Am wondering if you read my post as carefully as I read yours. I am pleased you have now acknowledged that sideways displacement of the sidestick gives a rate of roll, not an attitude (angle of bank). You seem also to have accepted that the white cross (that you call "caret") is not present in flight. (By the way, it is used - on the ground only - to give both pilots an indication of the position of the duty sidestick, and the resultant stick-to-surface deflections.)

We don't normally deal with the absolute basics of Airbus FBW philosophy on this thread, but to avoid misinformation this paragraph of yours had to be challenged:
"The A3XX sidestick (control) philosophy is not rate but attitude based, i.e. you move the sidestick to the side (and with it a caret on the PFD), and the bus will keep the attitude you demand by pointing that caret. So, for example, you move the caret in the right hand side of the PFD and the bus will maintain an attitude to follow that caret."

That was completely misleading, as I think you now realise. So, in your new post, you now write:
"While you initially give a rate based on stick displacement, the airplane then holds the commanded attitude (with centered controls!)."

That statement is correct. Good. But you continue:
"It hence does NOT behave like an airplane we all learnt to fly on, also the only airplanes we ever had any *real* upset recovery training in."

Also true, and I never said otherwise. Returning to your original post, your next paragraph reads:
"When in direct law [...], it'll start behaving like an airplane you learnt to fly on... and you're in effect flying a rate based control system..."

It is true to say that Direct Law is rather "like an airplane you learnt to fly on" (stick-to-control surface). But that is not "a rate-based control system", as I pointed out in my PS. So actually, the problem for Airbus FBW pilots is the transition from the rate-based Normal (or Alternate) Law, with which they have become accustomed in routine situations, to the (roughly speaking) stick-to-surface Direct Law. I don't think we yet know if the pilots ever had to handle that transition on this flight.

You now say: "...one should never implement systems which behave differently from what one learnt initially." While I have some sympathy with that point of view, I think in the real world it is a tall order when manufacturers are constantly striving to improve performance and efficiency.

Chris Scott

8th Apr 2011, 13:16

Quote from mm43:
I thought I had proposed a similar look at the relative forces involved in post #3106 (http://www.pprune.org/tech-log/395105-af-447-search-resume-156.html%20/%20post6354572), but may be it got overlooked, because I believe that Chris Scott came up with a similar scheme a little later in post #3143 (http://www.pprune.org/tech-log/395105-af-447-search-resume-158.html%20/%20post6356055).

It was also post #3134 (http://www.pprune.org/tech-log/395105-af-447-search-resume-157.html), actually, but you evidently got their first. You were just too subtle and succinct, I guess!

JD-EE, I think we agree about the initial sequence. CliveL has written an excellent post (http://www.pprune.org/tech-log/395105-af-447-search-resume-159.html), as did PJ2 (http://www.pprune.org/6356626-post3157.html), and they both seem to be on similar lines. It's what happens once the bogie/truck has fully emerged from the bay that is less clear.

Yesterday evening, exchanging PMs with HN39, he included the following : "I doubt that the difference in breaking force exerted at the two ends, integrated over the milliseconds time interval that it operates, against the moment of inertia of the gear in pivoting, would be sufficient to pivot the gear over 90 degrees to its 'down' position."

Taking account of what he had said, I offered him this: "I think the bogie/truck would barely decelerate initially, crashing through the flimsy door. But, as you say, it would soon be pivoting outwards (we have seen its pivot has survived). Also, by this time it would encounter the drag of the water. So I share your doubt that it would swing all the way into the fully extended position it looks to be in now. I should be very surprised if it had been in the extended position at the end of the flight, so further extension by gravity seems the most likely explanation."

Chris

auv-ee

8th Apr 2011, 14:21

Just posted to:

Sea Search Operations, phase 4 (http://www.bea.aero/en/enquetes/flight.af.447/sea.search.ops.phase.4.php)

Information, 8 April 2011
The ship Ile de Sein belonging to the company Alcatel-Lucent Submarine Networks (ASN), equipped with a ROV from Phoenix International Inc., is planned to undertake the fifth phase of maritime operations. This mission will be directed by the BEA and financed by the French State.

SaturnV

8th Apr 2011, 15:01

Ile de Sein is currently off the west coast of Africa heading north, probably to pick up equipment from Phoenix.

ILE DE SEIN - 9247039 - Vessel's Details and Current Position (http://www.marinetraffic.com/ais/shipdetails.aspx?MMSI=226235000)

Of perhaps more interest is this from the BEA:

The team on board the Alucia will complete vehicle operations on Friday. The vessel will leave the search area on Saturday 9 April and should reach the port of Suape (Brazil) on the morning of Tuesday 12 April.

This suggests either:
a.) WHOI is pausing its mapping and photography, and will resume after photos are interpreted, or when the Ile de Sein is enroute to the location;

b.) WHOI has photographed the recorders, or where they are likely to be, and thus there is no need for further photography at this time.

asc12

8th Apr 2011, 15:08

What a thorough description of the sonar imaging process and the creation of the images themselves. Thanks auv-ee.

When I first saw the sidescan image, I thought, "where are the remaining tail surfaces?" But, based on what you said, and my experience with radar, I'm not convinced that one should expect to see any tailplane segments in that image. It's tough enough to make out the wings.

I guess we'll have to wait for the Ile de Sein to learn any more, although I'd feel a little more comfortable if BEA would release some more images from the Woods Hole team.

bearfoil

8th Apr 2011, 15:09

Machaca - Pursuant to your excellent photos, I think it important to explain why some here are curious about inflight loss of VS. The two lug pairs and supports are from the front (1,2) of the attach, and show fracture due to rollover of the VS itself (side to side). The front lug has pulled with it the hoop structure attendant to the Aft Pressure bulkhead. The bulkhead is CFRP, and responsible for the integrity of the pressure vessel. If the VS separated at the outset of the loss of control, cabin pressure would be lost, and since the disc is not designed for rupture, it is reasonable to assume that some contents of the cabin may have spilled at altitude. This would describe some of the reports of flail and rupture to soft tissue evident in the victims. Of note also are the failure signatures of the rivets holding the lug bracket to the hoops. There is no longitudinal deflection of the hoop forward, which would be evident if the VS had rolled off the fuselage forward. In fact, the fractures suggest more than one episode of stress leading to complete failure. The composite Fin tabs have fractured in sideways aspect, and show no separation one from the other, a condition that would be expected had the single jolt theory of BEA occurred. A similar failure occurred on a 747 in Japan, leading to loss of the a/c due flight w/o Vertical Stabilizer.

Know that this is but a theory, and bears no malice toward aircraft or investigation, simply a recognition of upset leading to structure failure and loss.

HazelNuts39

8th Apr 2011, 16:05

bearfoil,

I fear you need to go back to BEA's Report No.2, and to the extensive discussions we have had earlier on this thread. The V/S has three pairs of attachments. The front pair stayed with the pressure bulkhead. What you are looking at in Machaca's first photo is a side view of the middle and rear attachments.

takata

8th Apr 2011, 16:20

No-one can so far explain why switching at 0210 to ALT2, due to unreliable airspeed, would cause an immediate upset as none of the similar events ever caused one! After being switched to manual control, any manual imput could cause one in relation with other unknown factors (workload, weather, technical, etc.)
Perhaps this is stated backwards.
"Why would an upset cause unreliable airspeed indications??", might be a better question.
The answer is in the link below and subsequent discussion.
AF 447 Search to resume

Note: With the discovery of the wreckage virtually under its 2010 position, the case for an immediate loss of control becomes stronger. The key question becomes how can it happen?

Well, sure, this question maybe stated backwards as, in my opinion, it can't be totaly ruled-out so far. Nonetheless, the A330 operational records should speak by itself as the probability for an "upset" happening some time before 0210, while flying in automatic (protected) mode at cruise level (A/P and A/THR ON) is clearly very very close to zero if related to her systems behavior.

One should remember that the only "system glitch" recorded after more than 15 years of operational duty accross a fleet of several hundred long haul aircraft was QF72. One will also note that this aircraft used a different (faulty?) hardware and would remember that this issue was not an "upset" properly: in fact, even if this situation could have caused some serious harm to the passengers, it did not cause an "upset" as the aircraft did not departed from its safe flight envelope (which doesn't mean that this problem was not dangerous for the aircraft safety).

Now, that the weather alone would put suddendly this aircraft close to an "upset" attitude, while flying in automatic mode at Mach 0.82, would certainly disconnect everything the same way. Then, if this was followed immediately by the freezing of all the probes, this would be quite a serious situation for the crew to manage. In this case, this would mean that they were not aware of the weather situation as they were not flying in "turbulence penetration mode", having their A/THR still ON up to this point. This would also be contrary to the right AF procedure and any cautious weather management at this dangerous spot.

Tim Vasquez' weather analysis was good but it pictured an aircraft position which was wrong as he took the projected "0215" position as reference alongside a wrong 0210 position.

concerning the wreckage position, I still doubt the 5 minutes "flat spin theory" direct from FL350 at a rate of 7,000 ft/mn while she was still powered all along her "free fall" (as no dual flameout occured). But who Knows?
The "deep stall theory", from the same level, will certainly imply a quite different aiframe (with T tail) and much more distance covered from LKP. The crash time (based on ACARS sendings) was estimated by the BEA between 0214:26 and 0215:14.

But, of course, it may have happened later in time, if she was not powered anymore after this point, as nothing, so far, can rule-out also this possibility.

We'll see what is found in the deep sea.

Machaca

8th Apr 2011, 16:34

bearfoil:
I think it important to explain why some here are curious about inflight loss of VS

That curiosity is sadly mistaken.

There is no way to explain the damage seen on the VS/lugs/hoops by any manuoevers and forces while airborne.

henra

8th Apr 2011, 17:10

This accident really remains a mystery.

I have to say I (along with some others) was strongly in favour of a search close to LKP. The really strange thing about it is now that they found it close to LKP I'm somehow at a loss as to why it is so close to LKP ?!

It would be really interesting to know where it is exactly, but if it is closer than 5nm from LKP, even a 'simple' spin scenario rather wouldn't do it.

From the Pulkovo crash we have a pretty good idea how long it can take from FL400 down to SL in a flat spin.
Looking at the Wing Loading of AF447 being ~580kg/m^2 compared to ~400kg/m^2 for the Pulkovo Tu154 we can assume that the vertical velocity /Rate of Descent in a flat spin should be a bit higher.
So the time for spinning from Fl400 all the way down should be below 200s, maybe rather 180s. (back then I roughly calculated terminal velocity of the A330 @210t as being in the order of magnitude of 140kts which seems reasonable with the ~120kts of the TU154 at Pulkovo))
Now let's assume the overall event took between 4,5 to 5 minutes. that would mean 90s - 120s before entry into the spin. if we decelerate from 480kts to 350kts the plane would travel between 10 and 15 minutes in that time span before entry into the spin. And even then there is still momentum to continue forward travel for some distance.

Looking at these assumptions I tend to assume there was some kind of course reversal, maybe unvoluntarily due to e.g. a massive wing drop ending up in opposite direction after recovery, maybe followed by a second upset this time losing it for good.
Unfortunately it really takes some creativity to get the plane to a point that close to LKP after a 5 minute struggle. Occam's razor seems to be off for vacation.

The other possibility would be that the speed was shed much quicker initially. With a full pull-up manouver it should be possible to decellerate to 350kts in maybe 30s. That would cover ~5nm. But then the remaining time of 4 - 4,5 minutes would require some form of recovery before the final plunge. All this w/o effectively traveling any distance.

None of the scenarios is really convincing. However, somehow it must have happened ????!

grity

8th Apr 2011, 17:19

http://i.dailymail.co.uk/i/pix/2009/07/16/article-1200132-05A74AF7000005DC-78_634x346.jpg

@bearfoil, this is a view from the bottom to the VS, it is nearly impossible to generate a nearly left/right symetricle break line like this in the air, every force to the VS in the air will generate a breake to the left or right and must end in a very asymmetric break line

a very nice front view to the gear of F-GZCP is this:
Photos: Airbus A330-203 Aircraft Pictures | Airliners.net (http://www.airliners.net/photo/Air-France/Airbus-A330-203/1864497/&sid=2761805c344402bbb48cd677d3587943)

takata

8th Apr 2011, 21:15

But does it matter if the gear was down. Surely the question is about what sequence of events took the a/c from level cruise to initial upset.
In some way, I guess it may give a hint about a stall recovery attempted but the crew, then a possible type of upset causing this crash.

All in all, what will certainly bring to us much more information are the ship from Alcatel-Lucent embarking the ROV from Phoenix International.
Here it is: L'Île de Sein, câblier (460 ft x 77 ft) , displacing 8,000 tons, built in 2002 by Hyundai Mipo Dockyard (South Korea) - her sister ships are: L'Île de Bréhat (at first named by BEA) & L'Île de Ré.
http://www.zdnet.fr/zdnet/i/edit/ne/2008/04/39380381/01-ile-de-sein-alcatel.jpg

promani

8th Apr 2011, 21:24

Mr Optimistic
"But does it matter if the gear was down. Surely the question is about what sequence of events took the a/c from level cruise to initial upset."

Well said my good man. We have had about 3/4 pages dedicated to landing gears. Next we will have posts concerning the toilet ACARS.

Takata, I have to say I read your posts with great enthusiasm, as you are so convincing in your statements, even if you might be wrong sometimes. You could convince me planet Earth was flat. :D

WilyB

8th Apr 2011, 22:15

The earth is NOT flat? :ooh:

mm43

8th Apr 2011, 22:36

By PM someone has asked me the following:-How will they find either the DFDR or CVR if they were torn from their mounts at impact?My simplistic answer, without knowing the exact distribution of the debris field, is locate the engines and compare the relationship to them of other parts of the aircraft, which will then provide an approximate heading at impact. Backtrack about 30 meters and start looking, or if unsure of the heading at impact, just start searching outward from the engines. If they are buried in the sand, magnetic anomaly detection could be used, or the treasure seekers inductive loop technique if using an ROV. The cable-laying vessel to be used in the Phase 5 operation - L'Île de Sein, will have all the necessary equipment, as dead cable detection is its "bread and butter" job.

ACLS65

8th Apr 2011, 22:36

promani
"We have had about 3/4 pages dedicated to landing gears. Next we will have posts concerning the toilet ACARS."

Let's see, if all the lavs were occupied, then the VS struck by lightning taking out the TCAS and weakening the always suspect VS, resulting in a mass IWE (Involuntary Waste Expulsion), the resulting synchronous flush, reduces the cabin pressure causing an apparent 300ft drop that triggers that ACARS msg, during the ensuing upset caused by the inability of the AB flight control system to deal with the situation (wouldn't happen on a Boeing mind you), the plane flies through the previously expelled falling fecal matter simultaneously blocking all three pitot tubes, the pilots drop the MLG to reduce the perceived overspeed... and so on.

Hopefully you all realize this is a humorous attempt to conflate many of the prevailing theories.

HarryMann

8th Apr 2011, 22:42

Looking at these assumptions I tend to assume there was some kind of course reversal, maybe unvoluntarily due to e.g. a massive wing drop ending up in opposite direction after recovery, maybe followed by a second upset this time losing it for good.
Unfortunately it really takes some creativity to get the plane to a point that close to LKP after a 5 minute struggle. Occam's razor seems to be off for vacation.

This is the sort of event chain that I seem to keep envisaging... Maybe because I've had a wing-drop stall (tip drag-stall, close to the ground), that having recovered, led (almost inevitably) to a repeat, and subsequent ground impact.

I find it strange that we are questioning the aircraft's return to near LKP, when we really have little or no idea of its flight path other than a mass and assumed vector at 02:10

I also view the seeming obsession with undercarriage position somewhat premature. Having postulated this (down position) and looked at any consequences in a few dozen posts, surely, there are then many more important preparations to be made for handling the information which is inevitably soon to be discovered and the ramifications it will have?

exosphere

8th Apr 2011, 22:42

I'm following this topic for a long time but I can't really find now at what altitude, speed and attitude the aircraft was at the time of the LKP. Reading the posts I can only assume it was in cruise flight at 350 at 0,82 and at normal attitude?

mm43

8th Apr 2011, 22:46

Originally posted by HazelNuts39 ...
It would be of interest to know the tolerance of the “credibility” test of the monitoring process internal to the TCAS which applies to the standard altitude parameter. It shouldn't be too difficult for the BEA to get that from the TCAS manufacturer.It took until the Report #2 for relevance of that ACARS message to be interpreted, which supprises me, on the basis that it appears to have been a non documented routine.

CONF iture

8th Apr 2011, 23:21

One will also note that this aircraft used a different (faulty?) hardware and would remember that this issue was not an "upset" properly: in fact, even if this situation could have caused some serious harm to the passengers, it did not cause an "upset" as the aircraft did not departed from its safe flight envelope (which doesn't mean that this problem was not dangerous for the aircraft safety).
The point on QF72 is that that airplane decided to do something on its own and on which the crew had absolutely no control.

In this case, this would mean that they were not aware of the weather situation as they were not flying in "turbulence penetration mode", having their A/THR still ON up to this point.
Unless thrust changes become excessive, it is not an Airbus request to disconnect A/THR for turbulence.

The NIST team has studied WTC7 quite thoroughly, and the study only confirms that the 9/11 conspiracy nuts are just that — conspiracy nuts.
'Thoroughly' until you read it.
I can only encourage people to read these guys (http://www.ae911truth.org/) too and evaluate if they qualify for the name of nuts ...

mm43

8th Apr 2011, 23:38

Originally posted by ventus45 ...
... there was no "post separation impact" with any other part of the aircraft, nothing other that water, and even that must have been "benign".Put it another way, "Why would there have not been any surface damage to the V/S if it had fluttered down from altitude?"

No, the benign landing in the water was IMHO a consequence of the fortuitous way that it parted company with the empennage at impact. A combination of a high CG and the noted fact that the tail was rotating to port at impact, allowed the V/S to depart and fall into clear water about 30 degrees to the left of the heading at impact. Its residual momentum probably allowed it then to drift clear of any other surface debris.

bubbers44

9th Apr 2011, 00:08

Probably more that the wind had an influence on it's movement with the current so moved it downwind of the bodies because it floated in the wind effect. Sort of like if you let loose of your boat in the water and the wind blew it away from you and you couldn't outswim it.

Machaca

9th Apr 2011, 00:47

ventus45 -- We're still waiting to hear what specific in-flight forces would cause the separation of AF447's VS.

Why the similar condition to the VS of ANZ A320 that hit the sea near Perpignan at 260+ knots?

ACLS65

9th Apr 2011, 01:14

More info on the Ile de Sein ship.

ILE DE SEIN - 9247039 - Vessel's Details and Current Position (http://www.marinetraffic.com/ais/shipdetails.aspx?mmsi=226235000)

Automatic Identification System (AIS) position and tracking.

Live Ships Map - AIS - Vessel Traffic and Positions (http://www.marinetraffic.com/ais/default.aspx?zoom=9&oldmmsi=226235000&olddate=4/5/2011%207:15:00%20PM)

http://photos.marinetraffic.com/ais/showphoto.aspx?photoid=252688
http://www.marinetraffic.com/ais/showallphotos.aspx?mmsi=226235000

takata

9th Apr 2011, 01:32

The point on QF72 is that that airplane decided to do something on its own and on which the crew had absolutely no control.
On the other hand, the fact that such incident happened only twice in the very same aircraft, as well as during the same flight, and that hundreds of other A330 flew safely many millions hours without ever experiencing it would give all the statistical depth necessary to conclude that this incident was obviously related to this particular aircraft and/or flight.

Unless thrust changes become excessive, it is not an Airbus request to disconnect A/THR for turbulence.
Unless I'm wrong, capitains are supposed to act like their Company is asking them to do by applying their Company published procedures.
Then, specific Air France procedure for A330 would ask for a speed reduced to M 0.80 and A/THR disconnected if an expected turbulence zone would be crossed. I don't think that they are different from Airbus ones in that respect.
Consequently, either this crew wasn't expecting any turbulence at all up to this point, either nobody was seriously looking at the weather radar.

I can only encourage people to read these guys (http://www.ae911truth.org/) too and evaluate if they qualify for the name of nuts ...
There is plenty of books published about the sociology of the rumor spreadings. Maybe you should try to read one and avoid to make a fool of yourself.

There was three official and independant "commission d'enquête", all made in order to convince the people from Abbeville that the flooding of the Somme area in May-June 2001 was not decided by the governement in Paris for protecting the capital from being flooded (even if the bassin of the Seine can't be pumped and deversed into the Somme). Nonetheless, most people from Abbeville are still believing the later today.

Turbine D

9th Apr 2011, 01:36

Originally posted by CONF iture...
...Unless thrust changes become excessive, it is not an Airbus request to disconnect A/THR for turbulence.

At 35 and .82 in A/THR while entering an area of disturbed weather, would the pilot choose to manually ease back on thrust to slow speed on the basis to make the Pax more comfortable?

I ask this question in that having passed through the first storm cell and suddenly entering the second cell, which was much more formidable, if speed was bled off intentionally and that this might have been the prelude to events that followed?:confused:

PJ2

9th Apr 2011, 01:47

TurbineD;

Turbulence penetration speed is 260kts at FL350, M0.78 higher, which would normally be selected on the FCU [Flight Control Unit on the glareshield]. Manual thrust is often selected to avoid large thrust changes with airspeed variation. The A330/A340 ALT mode is a "soft" hold which permits about a 50' altitude variation above/below the target altitude. This is primarily for fuel conservation purposes but permitting minor variations in turbulence also works. seem to recall that the books say that maintaining attitude and allowing altitude variations in severe conditions is preferred. Sounds easy, doesn't it?

On the VS...
I'm a bit surprised that the present discussion regarding the VS is taking on the dimensions of the MLG-extended discussion. It has been discussed at length in the past. I think it can be assumed that both will be resolved in the next few months.

That said, for those who argued two years ago and continue to do so today that the VS came off in flight, the question remains regarding the relatively pristine condition of most of the cabin interior parts, and, as we know now, the undersea wreckage.

We know from past accidents that a complete loss of the VS, as in this case, will, without fail in a swept-wing aircraft, result in a loss-of-control, almost certainly, but not inevitably, followed by a high-speed impact.

Notwithstanding that qualification, there isn't a single case of a transport category aircraft surviving the loss of the entire vertical stabilizer.*

We now know that the wreckage is concentrated indicating an intact aircraft at impact, but by the few photographs seen in public, large sections of the aircraft have survived relatively intact meaning a low-forward-speed impact. Contrast this with the condition of the Swissair 111 MD11 remains, even after a LOC from between 5000 and 2000ft, (TSB Report (http://www.tsb.gc.ca/eng/rapports-reports/aviation/1998/a98h0003/a98h0003.pdf), Sec 1.18.9.3, pg.196)**

The wreckage found on the surface, and now onsite, does not bear the earmarks of a high-forward-speed impact. I think a more reasonable conclusion is that the VS broke off at impact, ostensibly as mm43 describes above, (8th Apr 2011 16:38).

*I am fully aware of the photograph and survival of the Vietnam B52, which has been used to counter this example before. But some of the VS remains, enough, obviously to provide the lateral stability needed to make it home - I would be interested in an engineer's POV on this; the JapanAir B747 bulkhead accident remained aloft through extremely competent airmanship providing deft handling of power, but was eventually lost - the circumstances under which AF447 was operating would almost certainly preclude such a response.

**The TSB Report on SR111 is interesting reading for a number of reasons, some of which may possibly relate to a few aspects of AF447. Not thinking of a fire-on-board as a possibility but of comments regarding crew confusion from multiple warnings and system failures, and other aspects such as wreckage recovery, and investigation processes. Impact was 300kts, as explained in the report. The China Airlines B747-200 (http://www.asc.gov.tw/asc_en/accident_list_2.asp?accident_no=83) in-flight disintegration is also an interesting report to read in the light of both what is known and what may some expectations may be.

JD-EE

9th Apr 2011, 03:06

Sorry CONF iture, "I can only encourage people to read these guys (http://www.ae911truth.org/) too and evaluate if they qualify for the name of nuts ..." does not work. I have. Those idiots don't have a real clue about science, strength of materials, or the difficulty of keeping something secret.

(My hobby during my days at Rockwell International Anaheim was taking very vague hints of what was going on in the black world and noting things like what they were not talking about and telling them what they were doing and offering suggestions. It was fun freaking some of them out when I hit it right on the head. You CANNOT keep something that large a secret. By now somebody should have published information about size of charges, locations, who did it, and all the rest. Some idiot savant like me would have broken the whole thing with details FAR FAR less vague and hand-wavey than those cretins.)

JD-EE

9th Apr 2011, 03:14

PJ2, B2 and F-111 suggest that a plane can fly without a vertical stabilizer and swept back to the point of being triangular wings. If a computer can fly the plane a human could potentially learn how.

I suspect, however, that learning on the job in a crisis is not at all unlikely.

grity

9th Apr 2011, 04:52

For my money, "why the LACK of damage to the VS LE, it's sides, and rudder TE, not even scratches to the paint" has been the most compelling question from the release of photo number one the day it was spotted from the air, and confirmed with the divers recovering it. ventus, your money.....

there is a big damage on the rudder, like a shark bit, and if the chrash angel is more vertical than +/-45 degree the inertia energie let rotate the VS backwards to the blue sea

even the NZ paint on the VS in perpignan swam without scratches in the water, after stall............... with the same shark bit..... grity
http://static2.stuff.co.nz/1284666111/241/4138241.jpg

PJ2

9th Apr 2011, 04:59

JD-EE;

Yes, well, aircraft can be designed to fly without a vertical stabilizer when ways of controlling yaw are provided as in the cases cited. Also, as you'd know, some military aircraft require fbw because they are too unstable, (hence highly maneuverable) to hand fly. I believe "learning how" wouldn't be the impediment.

I agree with your last comment JD-EE because it's been proven a number of times with crippled transport aircraft. The results are mixed but that doesn't negate the likelihood.

In my estimation, under the circumstances which the AF447 crew faced, (I am assuming RP in the left seat, F/O in the right), re-learning how to fly an A330 with a missing VS, whether in ideal conditions or (depending upon when it is assumed by others that the VS departed before impact), with a stream of ECAM messages and audio warnings, in turbulence, in darkness, in cloud, with no airspeed information is an unlikely proposition and would result in the outcome described in my post above.

Machinbird

9th Apr 2011, 05:09

JD-EE, A moment of thread drift here:
"I can only encourage people to read these guys (http://www.ae911truth.org/) too and evaluate if they qualify for the name of nuts ..."They aren't nuts, but they are charlatans, and they are collecting money from the nuts they reel in for their cause. Their "explosives were used in the WTC " story is extremely flawed in that they have not accounted for the air trapped in the buildings as the upper floors came down like a monster piston.
Now back to AF447.

dvv

9th Apr 2011, 05:32

CONF iture, as I've read all the 9/11 NIST reports quite thoroughly, I can confidently say that all of them are very thorough, and none of the truthers' truthiness is anywhere close to the real scientific/engineering approach to facts and conclusions demonstrated by the NIST teams.

And oh, thanks for making my further reading of PPRuNe a bit more effective.

PJ2

9th Apr 2011, 07:04

CliveL;

Interesting regarding the possibility of flight without the VS, thank you. There is perhaps more controllability than I'm assuming?

However, I imagine your point regarding the hydraulic power seals the issue of controllability, as the rudder is powered by all three systems, (the THS by the blue and yellow) and a loss of fluid would be assured. That would leave engine thrust the only control for all three axes.

PJ2

HazelNuts39

9th Apr 2011, 09:13

It seems "loss of control" is now amply explained.

How about arm '36g'?

HazelNuts39

9th Apr 2011, 09:37

But that might allow enough airspeed pickup to allow the plane to get engines started and pull out to level flight. Using full rudder would help getting the nose down. You didn't forget to extend the flaps on the other wing to permit rolling it back to wings level?

JD-EE

9th Apr 2011, 10:24

HN39, you asked "Using full rudder would help getting the nose down. You didn't forget to extend the flaps on the other wing to permit rolling it back to wings level?"

Um, I'd rather envisioned retracting the flap that was temporarily extended if the plane gathers too much speed to safely have flaps extended. Regardless, the concept is that once it was broken out of its "spin" you'd want to return to a symmetrical configuration. Otherwise you simply buy yourself new problems.

john_tullamarine

9th Apr 2011, 10:57

http://www.pprune.org/tech-log/376433-af447.html links to a second locked thread on this subject.

HazelNuts39

9th Apr 2011, 11:03

Now, that the weather alone would put suddendly this aircraft close to an "upset" attitude, while flying in automatic mode at Mach 0.82, would certainly disconnect everything the same way. Then, if this was followed immediately by the freezing of all the probes, ... Any thoughts as to how (at FL350) "the weather alone would put suddenly this aircraft (in or) close to an "upset" attitude" in pitch, roll or yaw?

Regards,
HN39

PS1:: To illustrate what I have in mind, I have done a simple sum. In its Appendix 1 to BEA Report No.1, MétéoFrance writes:
The strongest vertical movements are observed in the "tower" of the cumulonimbus in its phase of rapid growth, that is to say before the top reaches the tropopause and the anvil is formed. The upward speeds can then reach 110 km/h and the downward speeds 50 km/h. The vertical speed can thus vary very rapidly inside of the cumulonimbus while crossing its "tower":
variations of more than 70 km/h in the space of 2 km have sometimes been observed. This intense turbulence can occur at the flight level of airliners and constitute a danger for them.As I read this, these are extremes, unlikely the fit the CB's encountered by AF447. But let's take them as written. An aircraft encountering an increasing updraft as described by MétéoFrance will pitch up, unless the pilot or AFCS opposes that with control inputs that maintain a set pitch attitude. The tail length of an A330 is about 29.25 m. Therefore the updraft velocity seen by the wing is 0.284 m/s greater than that seen by the tail. When the aircraft pitches up at 0.56 deg/sec, that difference will be nullified by the downward speed of the tail relative to the wing. At 483 kt TAS, the 'space of 2 km' is traversed in 7.45 sec, during which time the aircraft has pitched up 4.15 deg, less the lag attributable to its mass moment of inertia in pitch (which I have no data on).

PS2:: As a result of further research into the 'mathematical model' proposed in PS1 above, I should add that, due to several attenuating factors not taken into account in the 'simple sum', such as inertia, and the natural stability that induces a pitch-down moment when AoA increases due to an upward gust, the actual pitch up will be much less than that derived in PS1. Unfortunately, I do not have the more sophisticated 'mathematical model' that would produce the correct number.

JPI33600

9th Apr 2011, 14:09

First, a big thank you to all contributors for this fascinating and most interesting thread.

Anyway, I'm still puzzled by the (probable) crash site being so close to LKP, so I have used an online turn information calculator to sort a couple numbers, and here is how it comes.

Supposing AF447 was at cruise speed at LKP (around 500 kts), coming back close to this point in less than four minutes supposes a constant turn bank angle of 35° at constant speed.

If I reduce the speed to 410 kts, I still come with a constant turn bank angle of 30°. Any smaller turn bank angle doesn't allow reaching LKP again in a 4 mn interval.

Another hypothesis would be that AF447 reduced her speed considerably, which would allow reaching LKP again (after a 360° turn) in 4 mn using smaller bank angle values. However, I can't see any benefit in performing a 360° turn in the middle of the ocean, and any other evolution (e.g. a 180° turn to escape from bad weather) would involve still sharper bank angles.

As far as I can say, these high bank angle values may suggest that, apart from the final stall/spin/whatever event that ended in the crash itself, AF447 first experienced a massive enough "initial upset" that resulted in a change of direction that cannot be achieved using a standard rate turn. In other words, she may have rolled/tumbled heavily because of weather or anything else.

(Obviously, the speculations above suppose that the crash occurred at 02:14, as suggested by ACARS timeline.)

Assuming the scarce info we have is reliable, does anyone buy this theory ? :confused:

HazelNuts39

9th Apr 2011, 15:38

JPI33600;

Perhaps you would like to also look at a spiral dive. For 205t, FL350, M.82 the max. available loadfactor in normal law is n=1.662 corresponding to a bank angle of 53 degrees in a coördinated (no-slip) turn. The drag coëfficient in this condition results in a flight path angle for zero thrust of 6.4 degrees down, or a rate of descent of 5425 ft/min. Radius of turn is 2.6 nm, rate of turn 3 deg/s, so 360 deg. takes 2 minutes.

JPI33600

9th Apr 2011, 16:01

HN39

Thanks for your thorough answer. Regarding this spiral dive hypothesis, do you suggest an intended manoeuvre ? I suppose a 1.6 load factor spiral dive would certainly call for a "fasten seat belts" emergency announce, so most of pax and cabin crew would have been strapped ? Or are you thinking of an uncontrolled spiral dive (in this case, the cause for this dive would be as acceptable as anything else for what I called the "initial upset") ?

I may be wrong, but the more I think about it, the more I guess this plane has been initially affected by something huge and totally unexpected, with all unstrapped people being instantly killed or severely injured, and the direction of the plane being radically changed in a few seconds. A subsequent spiral dive followed by a (failed) recovery attempt would fit well in this kind of scenario.

HazelNuts39

9th Apr 2011, 16:20

JPI33600;

It's just Flight Mechanics expanding on yours. Hopefully the DFDR and CVR will tell us what happened and why.

BJ-ENG

9th Apr 2011, 17:29

@mm43
re your diagram of the aircraft impact (bg5kqh.jpg). I would be grateful if you could point me in the direction of the previous post it related to - date will do?

Chris Scott

9th Apr 2011, 18:18

Hi HN39,

Re the "could they return near to LKP in about 4mins?" issue, you write:
"Perhaps you would like to also look at a spiral dive. For 205t, FL350, M.82 the max. available loadfactor in normal law is n=1.662 corresponding to a bank angle of 53 degrees in a coördinated (no-slip) turn. The drag coëfficient in this condition results in a flight path angle for zero thrust of 6.4 degrees down, or a rate of descent of 5425 ft/min. Radius of turn is 2.6 nm, rate of turn 3 deg/s, so 360 deg. takes 2 minutes."

2 minutes allows quite a lot of "fat", so we don't even need as much as 53deg of bank. Can you or someone else confirm that deliberate bank angles beyond about 33deg (?) can only be sustained by maintaining the stick left or right of neutral throughout the turn, assuming Alternate Law (I forget which version they had) is in operation? Also that - beyond 33deg bank - the pitch-attitude would no longer be adjusted automatically to provide the vertical-axis 'g' (load-factor) appropriate to the bank angle, i.e., the aircraft would tend to fall out of the turn like a conventional aeroplane?

auv-ee

9th Apr 2011, 18:30

There is a new note from Jean-Paul Troadec posted on BEA's AF447, Phase-4 page. It summarizes the searches to-date on the occasion of terminating Phase 4, but, unfortunately, it provides no new information.

http://www.bea.aero/en/enquetes/flight.af.447/note.from.bea.director.end.phase4.pdf

mm43

9th Apr 2011, 18:32

BJ-ENG;

Here is the Impact Graphic (http://www.pprune.org/tech-log/395105-af-447-search-resume-104.html#post5902693) post, and you will also find related information a page or two before and after.

Machinbird

9th Apr 2011, 18:42

CliveL
However, what seems to be missing in this discussion - at least I have never seen it mentioned, is that if you lose the complete VS you leave all the hydraulic pipes supplying the rudder open to the atmosphere. Depending on how many of your hydraulic systems go up the VS you are likely to lose all hydraulic power and hence control in either axis.

Exactly. And none of the ACARS messages referred to hydraulic failure, and a hydraulic failure would definitely be the subject of a maintenance ACARS message.
Therefore the VS was on until water impact. Please, no more "They lost the VS before water impact." It is a lunatic theory.

mm43

9th Apr 2011, 18:59

Chris Scott;

A330/A340 - Flight Laws (http://countjustonce.com/a330/a330-flight-laws.html) , a HTML adaption of an original document by Andy Tracy may help. Note:: The summary table is also scrollable.

HazelNuts39

9th Apr 2011, 19:04

Can you or someone else confirm that deliberate bank angles beyond about 33deg (?) can only be sustained by maintaining the stick left or right of neutral throughout the turn, assuming Alternate Law (I forget which version they had) is in operation?

Hi Chris;

As I replied to JPI33600, it's basic flight mechanics, nothing more. I didn't look into the aspects you mention and expect that others, more qualified than I am in those things, will do that.

Rob21

9th Apr 2011, 19:31

According to brazilian press, a meeting will be held in Paris monday (11th) to discuss the next step, whitch is the recovery of the bodies (and wreackage?).

A person representing the families of the victims will be allowed to attend this meeting and also allowed to be an "observer in site" when the recovery of the bodies begins.

All bodies will be transported to France, where DNA examinations will occur.

gums

9th Apr 2011, 20:41

Salute!

Long time between posts, but frequent lurker. This the best site on the 'net for good poop and inputs from knowledgeable folks.

- I wish to point out to some folks that roll inputs that command a rate are not all that different from Champs, Cubs, Luscombes, P-51's, F-86's, F-4's and the Viper. Even if you must physically move the control that has direct/hydraulic connections to the ailerons or other surfaces such as I had in the F-102, you are not commanding a bank angle. Once you center the control or relax pressure you get whatever you are trimmed for, and the bent wing jets may continue to roll just a bit. Big deal. Unlike the Airbus, our Viper stick did not move at all until a year or so after initial operational jets were on the flightline, and thereafter about 1/8 inch to help feeling you had commanded max roll or gee command. No kidding. It was all pressure, and relaxing the grip stopped roll/pitch command instantly. About 17 pounds of pressure was max roll command to the computers. Biggest thing we had to get used to was "roll ratcheting", as the inertia of our arm woul reduce the pressure and the roll would slow or even stop, heh heh. Watch the T-birds versus the Blues. You will note that the Viper freezes after a roll and sometimes the Hornet will wobble a bit. In the Viper letting go of the stick by opening your hand stopped the roll instantly.

So the fact that the Airbus control moves is not a big deal. Only time I ever physically moved a stick for 30 years was in a violent maneuver, and not day-to-day flying, especially in close formation. I watched many great pilots in the T-33 and A-37 that had mechanically connected dual controls. Was very hard to tell where the stick was most of the time.

Thing that haunts me is the plethora of "control laws" that the Airbus employs. Shows the difference between the commercial planes and the ones I flew, I guess, but still don't understand the rationale.

- I would still not rule out the vertical stab coming off just prior to impact. Seem to remember the PIO on the Airbus at New York 10 years ago when the pilot exceeded the limits of the doofer. The recorders will tell us.

jcjeant

9th Apr 2011, 21:21

Hi,

Machinbird
Exactly. And none of the ACARS messages referred to hydraulic failure, and a hydraulic failure would definitely be the subject of a maintenance ACARS message.
Therefore the VS was on until water impact. Please, no more "They lost the VS before water impact." It is a lunatic theory. Do you know (are you sure) that the sending of ACARS message ended just a second before the aircraft made contact with water?
After the end of ACARS message ... who know if more was to be send .. but not send for any reason ?
Where in reality is the aircraft at the time of the end of ACARS message ?
Are that lunatic questions ?

HarryMann

9th Apr 2011, 23:04

So, jc, you're indicating you think that the VS came off quite a long way into the incident and thus was not an initiating nor early failure?

jcjeant

9th Apr 2011, 23:52

Hi,

Not at all :)
I just want to argue that if there were not any messages about hydraulic failure(s) (ACARS) .. we can not simply rule out with such argument the possibility of a hydraulic failure(s) after the last ACARS message received by AF maintenance. and so simply rule out the possibility of a VS problem
The aircraft is still flying after sending the last ACARS message or touches the sea he has now? (we know he is about 5.000 feet altitude cause last ACARS item send .... pressure difference )
And if he is still flying ... is that it can be prevented (or time) for any other reason to send ACARS ?

Exactly. And none of the ACARS messages referred to hydraulic failure, and a hydraulic failure would definitely be the subject of a maintenance ACARS message.
Therefore the VS was on until water impact. Please, no more "They lost the VS before water impact." It is a lunatic theory. My theory is that AF447 go out of its flight envelope .. because it is one of main risk when loss of speed measurement occur
I did not even read a report from BEA with this assertion .. but I think it will be written in its final report .. if the informations in the black boxes are still analyzable ....

Machinbird

10th Apr 2011, 01:29

2:14:26 WRN/WN0906010214 213100206ADVISORY CABIN VERTICAL SPEED

JCJeant
This was the last ACARS message. It indicates that an aircraft that had been at 35,000 feet 4.5 minutes before is now below ~8,000 feet and experiencing negative cabin pressure differential (If I remember the analysis correctly).

Some time after that the aircraft impacted the water at a high AOA and a relatively low forward speed indicative of a deep stall that was not an instantaneous deep stall, but instead indicative of a stall that had fully progressed into falling safe mode (because of the relatively low forward speed).

JCJ, are you hinting that the crew could have done something after this message that caused the VS to depart while airborne, or that the aircraft attitude created aerodynamic forces sufficient to break off the VS in a forward direction (a tail slide)?
Do you really think that there was sufficient time to do this, or energy available in the airstream at that point?
What about the arm 36g condition? How would you explain that?

BTW, (For those that believe the impact attitude resulted from a last minute stick snatch at the bottom of a flameout descent.) I very much doubt that the aircraft has enough elevator authority, in any control law, to pitch the nose up into a deep stall at the last instant before impact and still meet the encountered impact conditions. You would need a powered stabilator to do that, and the impact energy would be higher than that observed in AF447.

bearfoil

10th Apr 2011, 02:44

As with every excellent instructor, the professor supplies data to allow the students to come to an "independent" answer. Would have been honored to be a student of yours, PJ2.

bear

mm43

10th Apr 2011, 03:11

Regarding post #3244 on this page which has a link to the Impact Graphic page; the graphic should be read in conjunction with a revised rate of descent (http://www.pprune.org/tech-log/395105-af-447-search-resume-96.html#post5870418) which was recalculated following a timely intervention by HN39.

There will be some that reject the use of the ACARS message timings and the CPC Cabin Vertical Speed message in particular, but with little else to go on, and now that the aircraft is known to have crashed very close to its Last Known Position, the last ACARS message is relevant.

Reading the Impact Graphic posting (linked above), you will find reference to an earlier graphic which used half the velocities. In either case the vertical acceleration (ROD) at impact was high, as evidenced by the compact debris field. Arm 36g associated with the vertical stabilizer and rudder will probably have suffered the damage it did in either scenario.

I stand to be corrected.

CONF iture

10th Apr 2011, 04:06

On the other hand, the fact that such incident happened only twice in the very same aircraft, as well as during the same flight, and that hundreds of other A330 flew safely many millions hours without ever experiencing it would give all the statistical depth necessary to conclude that this incident was obviously related to this particular aircraft and/or flight.
Actually according to the ATSB another QF A330 has developed a similar pattern of fault messages. the crew completed the relevant procedures (introduced since the QF72 event) Did it allow to prevent a possible upset ? We don't know. Interesting to note that this special procedure, still in force to this day, does not guaranty to eliminate the risk but just to minimize the exposure time during which an eventual upset could develop.
Does it make the 330 an unsafe aircraft ? Absolutely not I do agree.

Then, specific Air France procedure for A330 would ask for a speed reduced to M 0.80 and A/THR disconnected if an expected turbulence zone would be crossed. I don't think that they are different from Airbus ones in that respect.
As stated in QRH 5.01, unless thrust changes become excessive, it is not an Airbus procedure to disconnect A/THR when going through turbulence.
It would be interesting if airfrance has decided not to follow Airbus on this one ... Would you have a reference ?

There is plenty of books published about the sociology of the rumor spreadings. Maybe you should try to read one and avoid to make a fool of yourself.
I'm OK with it.
Investigate building 7

JD-EE

10th Apr 2011, 04:12

jcjeant, methinks hydraulics failure would be a priority message. That strongly suggests that the VS was still attached up to one message time plus some Kentucky windage before the transmissions ended.

We have no real idea where the plane was at that time. The simplest suggestion is that transmissions ceased at the same time the plane ditched. The second simplest is a flameout for all engines, which, I believe, shuts down ACARS. So the plane could have still been a modest distance from its ditching location.

That aside, there is no compelling evidence that the VS was not attached at the time the plane met the ocean's surface and shattered. If someone postulates that a rudder setting somehow caused the VS to shear off they'd best be prepared to describe what the mount points, the clevis joints, that held the VS to the tail would look like in that case and show that they'd look like the damage we saw to the VS and the pieces of joint still attached.

The damage is particularly well suited to explanation via mm43's ditching diagram for the plane. The tail assembly hits the water with its nice wide flat surfaces and gets pushed upwards breaking at the bottom of the fuselage first. That leads to damage to the back end of the VS and rudder as well as a bunged nose on the leading edge where it was against the fuselage. The middle joints broke first then the others followed.

I cannot put together in my mind a scenario that has the VS mounting points come out looking the way we saw them in the pictures posted and in the BEA reports.

VS comes off in flight is not a scenario I can put together with the pictures and come out as a coherent picture.

Bobman84

10th Apr 2011, 06:12

All bodies will be transported to France, where DNA examinations will occur.

Hopefully the bodies themselves that are still inside the fuselage (as mentioned in the press) will yield vital clues as has happened in the past.

Would be nice to see more info about the search & recovery instead of 50 pages of vertical stabiliser nonsense ;)

HazelNuts39

10th Apr 2011, 09:03

my scenario has the VS separating effectively only a handfull of seconds prior to impact In that case, what significance does it have in the accident sequence?

considering the possible effects of resonant and flutter forces ?And no damage anywhere else on fin, rudder and hinges?

mm43

10th Apr 2011, 09:29

Originally posted by ventus45 ...
... the pilots, if still functioning on descent passing FL100, would, by that stage, have been open to trying "anything" to recover, including, full rudder inputs.But you have forgotten one thing, the RTLU ... it was found to be set at 7.9° +/- 0.1°, which is the limit in Normal Law at Mach 0.8 and FL350.... (272+/-2 KCAS).

2:11:21WRN/WN0906010210 272302006F/CTL RUD TRV LIM FAULT

Sorry, but I can't see sufficient loading from +/- 7.9° to achieve what you are suggesting, when the uninhibited maximum is +/-31.6°.

JD-EE

10th Apr 2011, 09:46

Loss of the VS is a lunatic theory
No, I do not see it as lunatic. I see it as unproductive to pursue because of its low probability and in your case relative unimportance.
I seem to have upset a few people.
No, you and many others replowing ground that is already quite well furrowed. We're not singling you out, even though I am not quite fisking your article. It just gets boring. Although I will admit that most people presume the VS departed the plane which started the whole mess.

The loss of the VS is a lunatic theory is it. So your collective assertion is that the VS on AF447 could not have separated in flight, and could only have separated at impact ?
I repeat, it is not lunatic. It is simply low probability given other facts as presented by BEA so far.

With all due respect to all of you, perhaps you have a different theory to the NTSB for AA587 ?
No. That is an entirely different event.

Note: - from the NTSB report:-
page 159
3. Conclusions
3.1 Findings
6. Flight 587’s vertical stabilizer performed in a manner that was consistent with its design and certification. The vertical stabilizer fractured from the fuselage in overstress, starting with the right rear lug while the vertical stabilizer was exposed to aerodynamic loads that were about twice the certified limit load design envelope and were more than the certified ultimate load design envelope.
Note the bolded text above. The shear from that event presented itself differently on the mounts than the AF447 damage we see in pictures. Just because the VS came off in both events does not mean they came off for the same reason. One must examine the damage to further diagnose the obvious fact that the VS and plane parted company.

Reference Document.
In-Flight Separation of Vertical Stabilizer American Airlines Flight 587
Airbus Industrie A300-605R, N14053
Belle Harbor, New York
November 12, 2001
Aircraft Accident Report: NTSB/AAR-04/04 PB2004-910404 Notation 7439B
http://www3.ntsb.gov/publictn/2004/AAR0404.pdf
Content References:
pages 31 to 42 inclusive.
pages 50 to 52 inclusive.
pages 55 to 60 inclusive.
page 159
3. Conclusions
3.1 Findings
6. Flight 587’s vertical stabilizer performed in a manner that was consistent with its design and certification. The vertical stabilizer fractured from the fuselage in overstress, starting with the right rear lug while the vertical stabilizer was exposed to aerodynamic loads that were about twice the certified limit load design envelope and were more than the certified ultimate load design envelope.
Did you intend this apparent repetition. It still contains the significant sentence fragment.

Now, I re-emphasise:-

"The vertical stabilizer fractured from the fuselage in overstress, starting with the right rear lug while the vertical stabilizer was exposed to aerodynamic loads that were about twice the certified limit load design envelope and were more than the certified ultimate load design envelope."
(sigh) Three times does not make it true, Lewis Carroll notwithstanding.
I did not say, and I am surprised that many of you have apparently assumed, that I meant, or was implying, that the VS separated at initial upset. I did not, and I do not.
OK, on that much we do agree.
I actually think the VS came off somewhere between 5,000 feet and 0 feet, but probably closer to 5,000 than zero.
You are free to think this. I depart from agreement here. It is not the simplest scenario.

At the risk of repeating myself from a prior message the evidence in hand strongly suggests that the VS was still attached until the last 15 seconds or so of ACARS transmissions. It MAY have come off at the time the last message was being transmitted.

It is also fairly apparent that the plane was below about 6k to 8k' at the time. That leaves room for your scenario. It would be incumbent upon you to describe what might have happened between 0 and 5k', picking your numbers, that would be significantly different as far as the VS is concerned than what happened at higher altitudes. Furthermore you should be prepared to describe how the VS assembly took on an appearance that it had been pushed forwards rather than sideways given the visible clevis joint related fractures.

It also strongly suggests the plane was out of control and falling out of the sky at the time. It less strongly suggests that flight control recovery had been effected and pull out proved to be impossible and in fact the plane died in the middle of a frantic pullout attempt. I happen to like that scenario as it explains a few things. But it's too thoroughly based on coincidence to make me happy.

My thinking is that there are many plausible ways that AF447's VS could be (and I think were) "exposed to aerodynamic loads that were about twice the certified limit load" (if you will - and more) when the aircraft entered the "denser air" at lower levels.
That is part of an explanation. What loads? What was the plane doing at the time that created huge loads?

To begin at TOD.

For various reasons, I belieive that the initial upset was an underspeed stall upset, which resulted in the aircraft entering a fully stalled descent, with a flight path angle about 30 degrees off vertical (60 degree dive) but in a fully stalled condition, basically in a slow flat spin, slowly yawing (my theory is happy either way, but, lets say to the right if we assume that the BEA is correct with their "tail was moving to the left at impact" idea).

That is a fairly likely scenario. But if they are only 10 miles away from LKP was it God's hand that appeared in front of the plane and slowed it fairly rapidly and then let it fall in a stall? Facetiousness aside, what slowed the plane down so suddenly? It wants to fly and in four minutes it was in the water from 35000'. Where did all the plane's energy go? And more importantly, why?

The tail was moving to the left idea comes from the way the pairs of clevis joints tore off the plane. Study the pictures of the recovered VS REALLY closely. Then play with holes punched in paper with a fitted dowel through the holes. Pull it off different ways and see what form the tears in the paper take. It's only a rough equivalence. But you should quickly see a pattern. You have to get the tail or at least part of the tail moving forward up over the fuselage in order to break it that way. And I agree, how in heaven's name did it avoid getting bunged up by other parts of the plane?

Possibly there was considerable surface winds and they picked up the tail as it broke lose and moved it the critical distance.

This mode of descent would be obviously unstable, but it would also have been unrecoverable, due to total stall of both mainplane and HS, and aft-ish CG. So it was probably oscillating slightly in all axis on the way down, ie, probably oscillating in pitch from about level to perhaps 20+ degrees up, in roll a bit, and in yaw rate a bit, but with yaw always positive. In other words, sort of like a fully stalled spiral descent to the right with a bit of Dutch Roll-ish type motion happening.

Now, air loads are a function of dynamic pressure, and I am damn sure that the pilots, if still functioning on descent passing FL100, would, by that stage, have been open to trying "anything" to recover, including, full rudder inputs. But before you jump to the false conclusion that I am saying "AA587 First Officer", I am not.

If the plane was making very low headway at the time it was within 5k' of the water AND rudder travel was limited to in flight high altitude deflection limits, how would full rudder provide enough force? It was going forward too slowly to remain airborne. So full rudder deflection would not be quite as exciting as it would near 300MPH at 35000'.

I am actually thinking more along the lines of turbulent eddies shedding off the fully stalled HS imparting latterally asymetric cyclic loads on the VS, (sort of like the FA18 LEX vortex busting on the fins idea - if it helps your mind's eye) such that, at some point, both the forces and the frequency were sufficient to either break outright, and or generate a resonant flutter like condition, that snapped it off at the base. The latterally asymetric cyclic loads would be generated by the varying wake fields from the port and starboard HS's due to the oscillating yaw rate. If you think it's out of left field, fine. I don't. Read the DSTO reports on the LEX vortex busts. Interesting stuff.

Perhaps. And maybe the recorders MIGHT be able to sort that event out from the impact. It seems to be a huge stretch with coincidences.

And, as someone else observed, why is it important? The important information is that information which leads to learning why this whole sequence happened in the first place.

Additional forces, either in or out of phase, could be generated by pilot rudder input. The combination of both, as dynamic pressures increased with decreasing altitude, may well have been "more than enough".

For those who assert that the rudder arm needed 36g to break is proof of impact forces, is it worth considering the possible effects of resonant and flutter forces ?

I think the posts asserting that there were no ACARS messages indicating hydraulic failures, "thus the VS was definately present at impact" is also flawed logic.

It does prove that the VS was still attached down to a small fraction of AF447's reported flight level. It was provably still on until at least a few milliseconds before the start of the last transmission. Otherwise the word that it had come off would have preempted other messages, I believe.

That does leave room for double flame out and a short glide from under the altitude at which the pressure warning would be issued. And it cannot be issued if the plane is gliding. That's a "falling" rapidly sort of message.

First, the ACARS stream ended well before impact (some seconds at least) did it not ?

We do not know. And the simplest presumption is that it ended coincident with impact. There was a hiatus in the transmissions that is suggestive of the plane going through some interesting gyrations. (Or perhaps had turned 180 degrees from its flight heading so the antenna was aimed as far away from the satellite as possible. I believe the antenna lobe is large enough that a bank would be required to take it off the satellite. I wish I knew the dynamics of the antenna aiming algorithms. I suspect they are fast enough to track through a tight turn. So that loss could mean a very severe bank.)

Second, even if they did only end at impact, and since my scenario has the VS separating effectively only a handfull of seconds prior to impact, would any hydraulic messages have been generated, prioritised, and injected into the ACARS transmit stream in time to have been actually transmitted in those couple of seconds ?

Please see above. It would take milliseconds for the message to be reformatted to note the massive hydraulics failure. So until a very short time before that last message started transmission the VS was there. We don't quite know where "there" was.

It could be that the VS was attached at impact, but I think those who assert with a passion that it definitely had to be, and was, are being a bit closed mind-ish.
It would be nice to know one way or another. But, the really important thing is "what started the avalanche?"