> >
>

# AF 447 Search to resume

Tech Log The very best in practical technical discussion on the web

# AF 447 Search to resume

14th Aug 2010, 23:38

Join Date: Jul 2009
Location: Not far from a big Lake
Age: 81
Posts: 1,454
Likes: 0
Received 0 Likes on 0 Posts
You can rule out fuel dumping guys. Very little of what is dumped down to 1000 feet will make it to the surface. By the time AF447 was at 1000 ft, it was seconds from impact.

Trying to be at maximum allowed fuel weight for landing aboard ship, we always tried to save some fuel for contingencies which we dumped at the last practical moment, often on downwind and turning into the groove. The stuff almost completly vaporizes within seconds. Of course, if you stick your canopy into the flight lead's visible fuel plume, you will get a film on the canopy. (Don't want it down the intakes though.)

15th Aug 2010, 01:23

Join Date: Jun 2009
Location: NNW of Antipodes
Age: 81
Posts: 1,330
Received 0 Likes on 0 Posts
HN39;

A misunderstanding of the sign of the pressure differential created the "error" in the Squair calculations! So the rational behind his argument fails.

However, the damage reported and the BEA's determination of how the flight ended do not match the implied vertical speed. A closer look at when the CVS advisory may have been initiated, and some juggling of a flight termination time could give a v/s that fits with the BEA reports (high vertical speed, 'en ligne de vol', and 'arm 36 g' etc..).

So, here is another look at what may have happened:-
Two independent pneumatic safety valves prevent the cabin pressure from going too high (8.85 psi above the external ambient pressure) or too low (- 1 psi below ambient pressure).

* hcbn = 7350 – 750 ft/min x 251/60 = 7350 ft - 3138 ft = 4212 ft

Now as a 2.03625437 inHg pressure change gives approximately a 2036 ft altitude change, and as the external pressure is 1 psi more than the cabin pressure (3), the external or aircraft altitude (hacft) is simply:

* hacft ? 4212 - 2036 ft = 2176 ft

2:14:15 - 2:14:29 = 14 sec > rate of descent = 9326 fpm *matches BEA summary of crash.

2:14:21 - 2:14:29 = 8 sec > rate of descent = 16320 fpm *excessive

The 9326 fpm rate uses the earliest time the Cabin Vertical Speed advisory could have originated, i.e. immediately following commencing transmission of the preceding fault report message with time of receipt 2:14:20. A further assumption is that the flight terminated 1 second after the Cabin Vertical Speed warning receipt confirmation by the aircraft [02:14:28].

NOTE: WRN messages take precedence over FLR messages.
2:14:20 FLR/FR0906010213 22833406AFS 1,,,,,,,FMGEC1(1CA1),INTERMITTENT
2:14:26 WRN/WN0906010214 213100206ADVISORY CABIN VERTICAL SPEED
Therefore, a vertical speed in the order of 9,000 fpm looks like a possibility with a crash time close to 02:14:30. I know this has been raised before, but with Svarin raising the issue, another look was worthwhile. In fact, I believe Machinbird offered a similar v/s some time ago.

mm43

Last edited by mm43; 15th Aug 2010 at 02:25. Reason: added "commencing transmission"

15th Aug 2010, 01:28

Join Date: Mar 2003
Location: BC
Age: 76
Posts: 2,484
Received 0 Likes on 0 Posts
mm43;

As a general rule, the 332 is fitted with fuel dump, the 333 is not.

That said, I agree with paull's and machinbird's comments.

Further, the jettison rate is approximately 1180kg/min. There was approximately 42.6k kg of fuel on board, - (BEA 1rst Interim Report, P.20) at LKP.

Even given a decision to dump fuel (which would have been both impractical (due time) and ineffective (there is no emergency in the QRH, and none I can think of that would require such action as an immediate need)), it is easy to calculate that there would have been a substantial amount of fuel on board at impact.

PJ2

15th Aug 2010, 02:21

Join Date: Jun 2009
Location: NNW of Antipodes
Age: 81
Posts: 1,330
Received 0 Likes on 0 Posts
PJ2;

I agree with your comments, and had previously said that a fuel dump wouldn't have left evidence on the surface to be detected by the SAR satellite radar. It has been suggested to me in a PM that an attempt could have been made to dump the Trim Tank (if it was even thought about!), though a transfer to the center tank would possibly have yielded a better CG result.

mm43

15th Aug 2010, 02:40

Join Date: Dec 2007
Location: Itinerant
Posts: 828
Received 76 Likes on 12 Posts
PJ2 (and others) have verified my contention that there is no scenario in which one would contemplate dumping fuel (if so equipped) in a mid-Atlantic ditching event. And even if fuel had been dumped at altitude (or otherwise released in a pre-impact break-up scenario) there is no way it would arrive at the surface, nor remain on the surface, in such a detectable form.

So let's agree to ignore the Drift Group's use of the phrase "voluntary or not" and simply deal with the "slick" or "pollution" or whatever it was that existed on the sea surface when the satellite took that shot. The "stain" (as I would prefer to call it because we don't know what it actually is) is approximately 1km in length at its longest point and perhaps half that in total width (widest point of separation of the two arms). Did it show in any other shots taken in the hours or days before or afterwards? Seems not, or that would have been reported, so all we know is that an anomaly existed that could have been kerosene at a time and in a location that generally corresponds with the disappearance of AF447.

So my point is simply this: If that exact location was not surveyed thoroughly by the undersea location resources when they were available, it could turn out to be a major blunder.

Discussion?

15th Aug 2010, 03:09

Join Date: Mar 2003
Location: BC
Age: 76
Posts: 2,484
Received 0 Likes on 0 Posts
mm43;
It has been suggested to me in a PM that an attempt could have been made to dump the Trim Tank (if it was even thought about!), though a transfer to the center tank would possibly have yielded a better CG result.
Yes - it's possible to dump the trim tank but indirectly: The jettison system removes fuel from the inner and center tanks which would in turn be "replenished" by the trim tank (valve opens on jettison) and outer tanks when the inners reach a certain level until a number of conditions are met which stop the dumping including crew control, the Jettison Gross Weight FMGEC entry.

The CG would move forward as you say. However, moving fuel forward from the trim tank to alter the CG, (the CG is thoroughly monitored by this and other systems and as such there shouldn't be a CG problem in any flight regime), can be accomplished without dumping if such is not required, (and given the dump rate, it is likely to be unhelpful in an emergency). The T.Tank selector switch can be used to OPEN the Trim Tank Transfer valve and all fuel will transfer forward to the center tank. The process of transfer seems to take a while but I've never directly observed/timed it.

In any case, I think we've kind of beat this one up...

PJ2

15th Aug 2010, 03:58

Join Date: Jun 2009
Location: NNW of Antipodes
Age: 81
Posts: 1,330
Received 0 Likes on 0 Posts
PJ2 wrote:-

In any case, I think we've kind of beat this one up...
... and a transfer let alone a dump wouldn't have been on anyone's mind.

Thanks
grizzled wrote:-

So my point is simply this: If that exact location was not surveyed thoroughly by the undersea location resources when they were available, it could turn out to be a major blunder.
In the cold light of day, the Drift Group were working with the positions of recovered debris, positions of drifter buoys that had passed by the area, and attempting to analyze satellite data on surface levels and winds to determine how those objects came to be where they were found. They have looked at the Pollution Spot (stain) and have determined that its position in relation to the other debris didn't make sense, and have legitimately dismissed it - though acknowledging its existence.

It may well prove to be a bad call, but we with 20/20 hindsight are left to argue and wonder about the spots relevance. In my original post on the subject, I indicated that further work may bring this Pollution Spot into sharper focus. On the other-hand it may turnout that the Drift Group were right. Either way, we are in the "know" and will be able to say, "We told you so!", or quietly merge into the background noise.

mm43

Last edited by mm43; 15th Aug 2010 at 05:20.

15th Aug 2010, 12:09

Join Date: May 2010
Location: PLanet Earth
Posts: 1,323
Received 103 Likes on 51 Posts
Regarding the pollution spot it would be important to know how frequent the satellite crossed/pictured that area, i.e. if it took images the days before and after.
Having that information one could conclude from the preceding or succeeding images the relevance of this pollution for AF447.

As a standalone information it is not sufficient to derive or rule out the relevance.

Any one here knowing/having an idea in which intervals this satellite would take pictures of the same area ?

But to be honest I'm a bit surprised BEA ruled out the relevance of this image/information so quickly and decidedly !?
Ruling out information that does not match one's preconceived idea of how thimgs happened is not a good idea in investigations of any kind.

And for sure not in the investigation of a still largely unexplained aircraft accident....

Maybe they have further information which allowed them to rule it out for good reason ?! But why not pass that information to the public then ???

15th Aug 2010, 12:20

Join Date: Jul 2009
Location: France - mostly
Age: 84
Posts: 1,682
Likes: 0
Received 0 Likes on 0 Posts
mm43;

thanks for pointing out the wider range of possibilities. I was just addressing the 'bottom line' of mr. Squair's analysis. So, based on the ACARS messages alone, the ROD in the final seconds could have been anywhere between 2463 and 16320 fpm, while the average ROD in the preceding descent was of the order of 7850 fpm: 35000 (2:10:10) - 2176 (2:14:21).

I also believe that the airplane probably stalled one or more times in those final minutes. Is it certain that it was stalled in the final seconds?

some juggling of a flight termination time could give a v/s that fits with the BEA reports (high vertical speed, 'en ligne de vol', and 'arm 36 g' etc..).
The 320 ditched in the Hudson at 780 fpm, 125 kt, 9.5 deg pitch. Its fuselage received substantial damage but stayed in one piece.

The Turkish 737 in Amsterdam impacted stalled at 3000 fpm, 100 kt, 22 deg. pitch, its fuselage broke into three sections.

AF447's stall speed (Vs1g; V_alphamax) was 158 kCAS; 162 kTAS; 274 fps. The corresponding alphamax can be estimated as 14.5 degrees. Interpreting "avec une assiette positive, une faible inclinaison et avec une vitesse verticale importante" as 2.5 degrees nose-up, leaves an FPA of 12 deg. down, or 3418 fpm.

So was it stalled?

regards,
HN39

Last edited by HazelNuts39; 16th Aug 2010 at 08:20. Reason: error corrected: Hudson ROD is 780 ft/min

15th Aug 2010, 16:23

Join Date: Jul 2009
Location: Not far from a big Lake
Age: 81
Posts: 1,454
Likes: 0
Received 0 Likes on 0 Posts
HN39
Just a gut feeling at the moment, but I suspect there is another still higher angle of attack where drag+ lift will balance the aircraft weight in a nearly level attitude. Problem is, no one has published the data for A330 Cd & Cl at ridiculously high angles of attack. You may be able to find data for other type jet transport accidents and get meaningful data since at very high angle of attack the actual shape of the airfoil becomes less important than wing area, aspect ratio, and wing sweep angle. In a 40 degree AOA for example, wing twist factors will likely have minimal impact. If sufficient data from older accidents is not available, I suppose you could use NASA panel codes to come up with meaningful data. Unfortunately I am not sufficiently versed in Aerodynamics to produce a useful result in a useful time frame.

15th Aug 2010, 16:37

Join Date: Jan 2008
Location: France
Age: 61
Posts: 9
Likes: 0
Received 0 Likes on 0 Posts
Fuel dumping

Just to stop the fuel dumping speculation, Air France A330s are not fitted with a fuel dumping system.

Last edited by pgroell; 15th Aug 2010 at 16:38. Reason: Typo

15th Aug 2010, 19:25
bearfoil
Guest

Posts: n/a
With great respect for all who post here, allow me to briefly summarize an area that has gotten little credence, perhaps understandably so.

The Industry, like others, is not without elements of other highly competitive and cutthroat endeavours at times. Betrayal is a part of modern commerce as it was in the days of the Silk Road and the Rum, molasses, and slave triangle.

Financial? Tech? Airframe? Regulatory Agencies, and investigative Boards?

Farming out highly critical aspects of a large and highly visible Investigation is par for the course, but comes with the possibility of failure (s), and inaccuracies. As the "Slick" was dismissed out of hand by BEA without evidence, followed by garbled and nonsensical language("Whether voluntary or not,") The fumbling of the Medical data, conclusions about structural evidence without "substantiating data", and the limitless hypothesizing about a flight path that is without evidence of any kind, What should have been a straightforward discussion of "Smoke on the Water" was ignored, actively, might I say, by the Trustees of Aviation safety, while 20 million was spent trying not to find the aircraft.

oily film? bunker fuel dumped by an anonymous ship,

Fuel? voluntary or not, which makes no Flight sense.

En Ligne de Vol? Useless phrase, impossible to know, and misleading, of course.

15th Aug 2010, 19:41

Join Date: Jun 2009
Location: NNW of Antipodes
Age: 81
Posts: 1,330
Received 0 Likes on 0 Posts
henra wrote:-

Any one here knowing/having an idea in which intervals this satellite would take pictures of the same area ?
The following COSMO satellite data may be of assistance. It is published by the Italian Space Agency, and even though some browsers may reject the site certificate, it appears to be "safe".

The SAR feature also has a SPOT mode which utilizes doppler/chirp technologies to provide a 1m by 1m resolution. Unfortunately, timely prior requests are required to enable scheduling of this feature over a particular area. As a matter of interest, the French representative organization for the COSMO project were represented on the Drift Group.

In 2009 there were 3 COSMO SAR capable satellites in polar sun synchronous orbit. I believe they were in the same orbital plane, equi-phased at 120° apart.

mm43

Last edited by mm43; 15th Aug 2010 at 19:52.

15th Aug 2010, 21:41

Join Date: Jul 2009
Location: France - mostly
Age: 84
Posts: 1,682
Likes: 0
Received 0 Likes on 0 Posts
Originally Posted by Machinbird
I suspect there is another still higher angle of attack where drag+ lift will balance the aircraft weight in a nearly level attitude.
Yes, your quite right, but it requires a higher airspeed. The lift-coefficient at alpha-max is 1.45 at sealevel. The proverbial barn door held perpendicular to the airstream has a drag-coefficient of the order of 1.1. I would think that your condition is somewhere in between, but suspect that it is closer to the barn door than to a high-tech wing with attached flow.

regards,
HN39

PS:: For example, taking the coefficient 1.1 for the resultant force of lift and drag (ignoring thrust); your AoA of 40 deg, and mm43's pitch of 5 deg. NU, would give a rate of descent of 10840 ft/min at 181 kCAS.

Last edited by HazelNuts39; 16th Aug 2010 at 17:01. Reason: PS

15th Aug 2010, 22:54

Join Date: Nov 2001
Location: UK
Age: 70
Posts: 76
Likes: 0
Received 0 Likes on 0 Posts
The thing I find interesting about the Cosmo image was it was taken 30 hours after the event before any wreckage had been sighted by the search teams.
Now, understandably that satellite probably downloads thousands of images which there is little point in analysing, but it’s curious this one may have languished on a server until someone had a brainwave and checked it.

Wonder when that was.

15th Aug 2010, 22:57

Join Date: Jun 2009
Location: NNW of Antipodes
Age: 81
Posts: 1,330
Received 0 Likes on 0 Posts
HN39 wrote:-

I also believe that the airplane probably stalled one or more times in those final minutes. Is it certain that it was stalled in the final seconds?
That is relevant, and is the question that somehow doesn't seem to generate a positive answer, especially when we are faced with -
"avec une assiette positive, une faible inclinaison et avec une vitesse verticale importante"
or - "with a positive attitude, a low [bank] angle and with a significant vertical velocity"

BEA Reports provide some indicators that support the stall theory, but whether the descent was marked by a series of stalls and partial recoveries, is like everything we are dealing with, just conjecture. The tail yawing to port on impact provides (imo) a small clue as to how the "positive attitude", "low bank angle" and "vertical speed" were attained, i.e. in a "deep stall" where a clockwise rotation had a stabilizing outcome in the "lost lift" vortex.

The ROD from your calculations with 5 degrees nose-up may be interesting.

Damage to the Radome showed a significant upward moment, but the horizontal moment at impact failed to shatter the top section. So the GS at impact "en ligne de vol" would appear to be quite low - 60 knots?

mm43

15th Aug 2010, 23:03

Join Date: Jun 2009
Location: NNW of Antipodes
Age: 81
Posts: 1,330
Received 0 Likes on 0 Posts
BackOffice;

At a guess, not until the Drift Group was set up, and those with the appropriate expertise knew where to go "fishing".

mm43

16th Aug 2010, 00:22

Join Date: Jan 2008
Location: Herts, UK
Posts: 748
Likes: 0
Received 0 Likes on 0 Posts
The 320 ditched in the Hudson at 1500 fpm. Its fuselage received substantial damage but stayed in one piece.
25 ft/s at touchdown, during and after the flare... ?

Last edited by HarryMann; 16th Aug 2010 at 11:00.

16th Aug 2010, 01:29

Join Date: May 2010
Location: Le Shed on the Tropic of Capricorn
Age: 62
Posts: 22
Likes: 0
Received 0 Likes on 0 Posts
Drift re-analysed using a simple approach

The Drift Group did their analysis by calibrating their models to things like the drift buoys etc which were drogued relatively shallow compared to a deep ocean but relatively deep when compared to floating debris. If we just take the first 22 bodies (due to their improved consistency relative to other items and not too much time from the impact), calibrate the model to the pollution spot (or slick) and using a method of least squares (Excel) force the curves to go through this point then the potential point of impact extrapolates to 30.51W 2.62N.
(This is achieved by changing time to slick time and using a grid centred on the slick to force the curve through this origin, then converting back from slick time and slick co-ordinates ie using a slick method). The items in the analysis that appear furthest from the ‘true body line’ (when all items are considered) are the first sightings by the merchant vessel then by the navy. Also the VS recovery position and then (to a lesser extent) its sighting position. The first 2 bodies are also slightly north of the ‘expectation’.
These bodies show a spreading out in an ever increasing wide ‘triangle’ (over this180 hours) from the extrapolated impact focal point as they head north which would be expected due to ‘randomness’. There is also a very slight ‘general’ easterly drift. The bodies are found with very little northerly spread (at one time period) but both this northerly and the easterly trends may not be statistically relevant and may simply be the way the retrieval was done considering only 10% of the bodies were recovered.
So IMO the slick is statistically relevant and its position is useful.
regards. Ian

16th Aug 2010, 02:59

Join Date: Dec 2007
Location: Itinerant
Posts: 828
Received 76 Likes on 12 Posts
FluidFlow:

Wow. Thanks for that work and contribution.

mm43: In light of all your incredible and insightful work I'd like very much to hear your comments on Ian's post.

grizz

Last edited by grizzled; 16th Aug 2010 at 03:18.