PRIM 1 and SEC 1 Faults
 

Quote from Svarin (Apr12/1754z, currently #3387 on Page 170):
"02:13:45 F/CTL PRIM1 FAULT
02:13:51 F/CTL SEC1 FAULT
First, this describes a simultaneous failure of PRIM1 and SEC1, which greatly reduces the likelyhood of a manual shutdown or reset."

Your analyses in the area of possible wiring fault and possible pitch-up have been most thought-provoking. Thanks.

However, I'm not entirely convinced the two faults were necessarily (a) simultaneous and (b) "permanent" (for want of a better word).

Re my (a), I agree that SEC1 is probably below PRIM 1 in CMC transmission priority, so they might have been simultaneous. But could there not equally be a gap of up to (say) 11 seconds?

Re my (b), these are cockpit (ECAM) warnings that have sufficient maintenance implication to justify an ACARS message. The BEA comments, in both cases: "this shutdown could be the result of a command or of a failure."

Am wondering if the crew might have gone for a quick trip-and-reset (re-boot) on PRIM 1, followed almost immediately (not simultaneously) by the same action on SEC 1. (We used sometimes to resort to similar expedients in the early days on the A320. Don't know what current thinking on the A330 is, let alone SOPs – but they would have been out of the window by 0213z.)

If a computer had not displayed a FAULT W/L (warning light) prior to a crew-selected re-boot, the fault W/L and associated ECAM message might last for just a few seconds. The question is, would it subsequently be transmitted by ACARS? The other, remote possibility is that the crew might have left one or both PBs (push-buttons) in the OFF position, deliberately or otherwise.


CONF iture,
I agree with your above point that speed reduction to turbulence speed might have been in progress at 0210z, implying low N1. Just a reminder that the CF6 engine does not use EPR, N1 being the primary indication of thrust.

Chris

By FCOM and QRH, a PRIM or SEC reset (one at a time !) could be attemted by a crew if a failure is suspected or detected.
From my experience on the 330, I have never had or thought to reset such computer unless specifically requested by ECAM or QRH.

I take note for the CF6. Thanks.

CONF iture;
Yes, fully agree - it is a serious matter that requires some careful thought and action.
Selecting the turbulence penetration speed on the FCU would not reduce the N1 into a low range.

The cruise N1 for M0.82 would be around 86%. The turbulence penetration speed at FL350 is 260kts, (M0.78 @ FL370) and the power setting is 76%.

Setting 260kts in the FCU speed window may initially reduce the N1 below the 76% and as the speed was achieved would return to the target setting but any reduction would not be in a 'low range', nor would a disconnection (caused by system problems) leave the aircraft at risk.

Sometimes in mountain wave conditions where the speed is increasing, the A/T will reduce the N1, sometimes even to idle if the Mach is getting close to VMO/MMO but for a twenty to twenty-five knot reduction the N1 would not reduce into what could be called "low range". Even if it actually did, it takes a long time to lose enough speed to be at risk.

The thrust lock function is easily handled using manual thrust, (move the levers to set 76%). The severe turbulence penetration drill states that the A/T should be disconnected if speed variations are large.

Just a quick question. (Apols if already been posted)

Is there any way the reverse thrusters could have been deployed with all the apparent failings going on the flight deck, or is this outcome totally impossible due to fail safes. (if so, would you get a record of this in ACARS?)

Thanks.

CONF iture and PJ2,

Thanks for the information on the SOPs and practcalities of FCC resets. As I said:
"but they [SOPs] would have been out of the window by 0213z"

Perhaps I should have said "might". Aside from SOPs, sad to say, there was little time left for "careful thought and action".

It is a tantalising, if understandable, feature of these WRN ("cockpit effect messages") from ACARS that their duration seems to be indeterminate.

Not totally impossible, but.................VERY UNLIKE.

I agree, and that's what we encounter most of the time when confronted to a low N1. Nevertheless, an updraft due to convective weather could initiate a similar N1 reduction, and so even if the speed was already established at M0.80

Note that the turbulence penetration speed for the 332 is M0.80 @ FL370 compared to M0.78 for the 333 I believe.

Chris, CONF iture;

Agree that things unfolded quickly and that time to consider and act was short. I think we're bumping up against the limitations of text and typing...we've flown the Airbus so I think we're meaning and understanding in the same way. In my post I just wanted to clarify that the N1s would not be in a low range regime as a result of selecting a turbulence speed even if the a/t disconnected right at that point. I'm assuming here that the idea behind the low N1 range was, the aircraft would lose speed and stall given that it was in Alt2 law, etc? ...if not, then I haven't, (not for the first time!), understood.

The notion behind WRG as I understood it from some reading and research was that it was not necessarily a physical "wiring" matter, but was a term used to describe a situation where the two FWCs could not independently interrogate the computers sending the original fault message(s) such that the "correlation" stage of fault determination could complete. The reason (again making great assumptions and interpretations here as I am not an engineer), that perhaps the timings of the message was not long enough for the process to complete.

Again, bearing in mind that the purpose of the [Computer fault] > FWS/FWC> CMC > ACARS process, (aside from warning the crew, etc), is to provide maintenance with sufficient information to understand the fault well enough to prepare for handling it at arrival, (ostensibly to avoid delays, or expensive repairs which don't fix the problem, etc).

As we're keenly aware, we're using the ACARS data in the same way investigators would use flight recorder information but with 26 text messages instead of 1800 or so aircraft and flight parameters and the voice/sound recordings. We're looking at minute entrails to see what produced the particular combination of messages in (roughly) the timings seen. The limitations are severe even with the photographs but I'll be surprised if, throughout the thread, some haven't come very close to what happened.

CONF iture, thanks for keeping me honest! Are you on the A330 now?

HazleNuts

I was not aware of there being a cyclic nature to the high speed buffet. The only times I have experienced the high speed buffet on all three types I did not experience that but, in general, feel that it was less 'rough' and of a higher frequency than the low speed buffet. But it was all a long time ago and memory plays tricks.

Obviouslty, too, my experience was on older aircraft not FBW, and neither do I know what stall training is done these days. However, if it is only to the onset of buffet or stall warning, pilots will not know the stall characteristics of their aircraft. Also, if it is done using the traditional method of trimming to 1.3Vs and then reducing airspeed at 1kt/sec they will be unaware of two important factors:-

(1) In an unintentional stall following an unnoticed decay of airspeed (Turkish Airlines at Amsterdam is a case in point) the autopilot will have trimmed the stabiliser nose up. Therefore, when the pilot belatedly recognises the situation, the aircraft will be severely out of trim with, perhaps, insufficient elevator authority to accomplsh a straight forward recovery, and this will be excerbated with underslung wing mounted engines when power is applied.

(2) In the case of an accelerated stall caused, either by pulling G or by a large gust or both, the time between the onset of buffet and the stall itself may be very short, only 1 or 2 seconds - insufficient time for most pilots to recognise the problem.

Of course, if what others have said about the Airbus aircraft shouting 'stall' 'stall' 'stall', stall recognition should be obvious! However, when you have lost envelope protection (and here again I do not know enough about FBW), and with modern wings designed for max efficiency in cruise is there a tendancy in some conditions for the aircraft to continue pitching nose up? If there is, and AF447 was in alternate law, it might be possible to get into a super stall very quickly.

I won't add more to the already extensive speculations in this thread, as other have said let's wait for the recorders. But I do wonder whether enough training is done on all aspects of the stall rather than just demonstrating the approach to and recovery from the classic stall.

For the last 12 years actually, PJ2, but of course still learning a LOT. Too bad I seem to forget at a similar rate if not more …

"Too bad I seem to forget at a similar rate if not more …"

...tell me about it...

I was using 333 information, but now have the 332.

Hi,

Is it possible the "vertical compression forces from bottom" seen by BEA (and supposed due to deceleration (~100g) with a/c quite horizontal) originates, from water pressure, with a/c going meters underwater (nose down descent all the way)? (argument: engines "peeled off").

Would be "Dozens of bodies are present among the wreckage." ( Le Figaro - France : Le repchage des corps des victimes en question ) preventing BEA for showing pictures of the tail?

Sorry for my very rude English,

@Mr Optimistic, in basic arodynamic for every flight part angle between the best gliding-angle and 90 degree (vertical) exist one relation between lift and drag and one corosponding AOA for this angle,

and it results in a relativ stable decending speed, depending on wing area/profile/wight/place of CG/etc so it must not need the steering of the hand of god to fall from 10000m within a few km to nearly every place into the water

look at the polar diagram of mr.lilienthal
The aerodynamics of Lilienthal

so who know which AOA result in which decending speed, and which place of CG (or which elevator pitch) make this way down then stable....
grity

Speaking of AoA, I finally found something I think was raised in previous discussions (search did me no good in finding the thread)

That's a lot of if, however as this bulletin was issued in 2010, it may be that the AF 447 mishap was related to discovering discrepancies.

And it may also be unrelated. I note that it would take multiple failures, and multiple probes well past manufacture date, and in service, having this problem all at once for it to have influenced AF 447.

Back at post #3478 , swordfish41 made reference to the sidescan sonar nadir/shadow made by a large piece of wreckage to the east of the debris field. An enlarged sidescan sonar image of the debris field published by the BEA is reproduced below.

http://i53.tinypic.com/2ceix60.jpg

I looked at the sidescan image a couple of days ago and enlarged it to discover the curved shadow, and suspect that it is the forward fuselage. The length would seem correct, but the skin may have been split away from the frames to present an enlarged width. I have good reason to suspect that swordfish41's analysis is reliable, as he was responsible for the finding of the HMS "Ark Royal", that sank in the Mediterranean in 1941, by using 200kHz sidescan sonar for a BBC documentary. I posted some information on the finding of the Ark Royal in post #1038. A post in response by swordfish41 is at post #1049.

With regard to the AF447 wreckage, the forward fuselage probably parted company with the rest at impact, and with less impediment to its forward movement, made its way to the bottom on a slightly different trajectory than other debris. If this is the case, then the heading at impact was around 070°T.

EDIT :: True north is 10° to the top right of the Y axis in the above graphic.

grity; BJ-ENG;

The impact graphics at post #2066, seem to match the injuries resulting from a 100g+/- impact as researched by BJ-ENG. Also, grity's analysis of the likey 'g' forces tends to match what I had proposed, and the small debris field found would tend to support that. Thanks guys.

Water pressure does not affect people that way. The only cavities in the body are the lungs, sinuses and the windpipe between them. All other parts are water filled and will equalize to the ambient pressure without significant dimensional change. The record scuba dive (no pressure protection) is to 330m. (Deep diving - Wikipedia, the free encyclopedia) Pressure on a person with un-equalized cavities (person not breathing from ambient pressure source), will cause tissue damage and bleeding from broken blood vessels near the cavities, but not breaking of bones. (However, the free dive - breath-hold - record is 214m (Free-diving - Wikipedia, the free encyclopedia), so with training, even breath-hold is not much of a limit.)

That could be one reason.

Nothing wrong with your English.

mm43

Thanks for the enlargement. I had looked at this photo (not enlarged) and wondered if it (the little separate blip) might be what is left of the tail section. There is a fair amount of mass back there and it may have settled on a different trajectory as you pointed out. I say this as there has been no mention of the forward fuselage, but speculation (confirmed or not) of finding the tail section. Obviously, there are no doubt additional photos of the wreckage, which have not been released for various reasons, but exist in the hand of the investigating authorities and they know what this sidescan image represents. I suppose we will know soon.

Yes Grity thanks, good way of looking at it. We'll assume up to circa 100g, anywhere form 60 to 100 ++ seems the ball park actually. Certainly fits with some of the post mortem findings .

mm43 and swordfish41,

Is there a possibility that Turbine D is right, and that it is the tail section (by which I think he is referring to the aft fuselage)?

Would the debris field alignment (approximately 070/250) necessarily be roughly indicative of the heading at sea-level impact, despite currents on the way down? If the curved piece in the north-east was the aft fuselage, would that suggest the Hdg 250 rather than the 070?

swordfish41,

Do you think the engines may be present in this pictured debris field? Could the laterally-lying pieces extending northwards include a wing box?

Could be. The fly-by height and resolution of the sidescan sonar is not helping, and trying to piece together what other items could be, would be stretching it a bit far. The piece in question has got some height to it, and hence the shadow.

My interpretation of the BEA's press release, was that what was shown on the scan was the extent of the debris field. I believe everything that went to the bottom is represented there.

CS, mm43, and perhaps swordfish41,

With all due respect, I think you may be trying to read too much into a few low resolution blips in a sonar image. The dark spot does look like a shadow, but if it is, we should do the math for it.

Using the provided scale:
The eastern target is 360m from the (eastern) vehicle track.
The length of the shadow is 11-13m long.

The vehicle is typically flown at 10% of the maximum range setting of the sonar (10% gives a good compromise between pronounced shadows and not letting the grazing angle on the seabed get too shallow for receiving good back scatter). For the 600m range used here, the altitude of the vehicle is about 60m.

That puts the target 6 times as far from the vehicle as its altitude. Thus the shadow is 6 times longer than the height of the object casting it. So the object is standing about 2m above the sea floor. Does that match your other assumptions? (I see from our intersecting posts that mm43 is wondering, too.)

I agree with you totally, as per my post above, it could be a stretch too far!

The only thing that the target in question has is a bit of height that has been amplified by the grazing angle of the sonar.

Re the magnified scan, mm43, the wings seem to be awol. I'd expect SOME indication in the distribution that the wings were down there with the plane, somewhere. I'd expect them to be at least a slightly denser collection of debris in a roughly straight line perpendicular to the rest of the debris.

That outlier might simply be the outer portion of a wing if it broke off on impact. That would suggest the other wing might be that slightly thicker bit of debris across the roughly fuselage shaped debris field.

If that is the case we might even be able to infer something modest about the current distribution as the plane sank.

I think that is mainly what is to be learned. (I'd love to be wrong about that.) It would be interesting to rework the current data provided by Hyperveloce way back here: http://www.pprune.org/5693490-post1035.html

I meant to look at that a few nights ago but could not find the data. By coincidence, following a link to swordfish41's comments about the Ark Royal lead me right to that discussion of currents.

Scale
 

These two images show the approximate relative sizes of the debris field and the aircraft.
--------

http://pickyperkins.home.infionline.net/bigfield.jpg

--------

http://pickyperkins.home.infionline.net/rhfield.jpg
--------
The red "pixel" near the aircraft in the second image represents what I imagine to be the effective sonar resolution, based on my understanding (or mis-understanding, as the case may be) of auv-ee's post #3169, "Resolution of sonar images", describing how sonar side-scan images are constructed. The red "pixel" is about 6 m from top to bottom, and the width arbitrally narrower. From the auv-ee's description I understand that the left-to-right resolution can be made as narrow as the constructor wants, but from the image I am guessing not very narrow, to conserve memory requirements. (auv-ee please correct this as necessary.)
---
http://pickyperkins.home.infionline.net/pi.gif
---

Scale
 

PickyPerkins: Very nice, thank you. I rest my case.

Chris Scott wrote:
I think auv-ee addressed this recently - he suggests it's probably more indicative of the underwater currents. I think it also depends on the degree of fragmentation. We also don't know if this is the entire field (though it looks to be).

My understanding[*] is that if parts are relatively intact, you get a sort of ballistic trajectory that depends on the resistance offered by the falling parts. You can see an indication of this in the HMAS Sydney report -- see the final page of:

http://www.defence.gov.au/sydneyii/f...apter%2016.pdf

and again on p 218 of

http://www.defence.gov.au/sydneyii/f...apter%2012.pdf

The relative distribution of the parts in turn will be affected by underwater currents pushing on them during the time of descent, but unless they further fragment (eg implosion) what you see in the sonar tends to reflect their original shape.

The question of the relation between resting orientation and orientation at impact is something you'd need expert comment on. My lay impression is that a relatively intact airframe would tend to fall through water nose first, as I presume that would offer least resistance.

However, if the wreckage is heavily fragmented, my understanding is that the final debris field is often spread much further, because each individual fragment behaves differently. The pieces which are both densest and offer least resistance, descend fastest and move least with current. Pieces that are lightest and with the most resistance take the longest to fall and are pushed furthest by the current. Other things fall in between.

The sidescan suggests large parts could be contiguous, but the few photos do show a lot of fragmentation of parts - see at Oprations de recherche en mer : images du site , especially the one titled "Partie de fuselage".

It's hard to tell without seeing more of the cabin, but I assume it hasn't been shown because human remains were visible. From mm43's post, it does however look like the main part of the debris field is about 400m long out of the 600m shown, whereas the intact A332 aircraft is about 59m long? Given the amount of internal parts (eg galley parts) found on the surface, from both galleys, I assume there were either multiple breaks in the fuselage, and/or rupturing, to allow them to escape.

nb[*] Several people here have already shown that they have much more knowledge than I on this and I would defer to them if the above is wrong.

Also, compare the tidiness of the HMAS Sydney image with the AF 447 one. I don't know enough to know if this means anything.

@auv-ee: thank you for your documented answer #3504.

About bodies injuries vs a/c attitude: it seems (imho) that there is kind of circular arguments (tautology) about them:So, it is fully compatible with the a/c attitude supposed by BEA.
BUT. Bodies recovered are probably those from passengers who were not attached. And they could have been in every position and place in the cabin at impact, depending of what happened before.

In fact, I can't imagine this aircraft falling from 35000' to 0 keeping a "normal" attitude all along (and not attached passengers remaining in their seats*). Adding a storm, is this "flat" fall truly possible ?


*If that was which happened, thanks for it: passengers kept sleeping all the way. It had to be more rude for the crew.

Forgive the noob question on sonar, but is the image produced on a single pass, eg from the rov above the eastern edge of the lane, or is it the addition of two half pictures, one from the east and one from the western pass ? Also what would the effective lateral beam width be, ie the length of image producted by a point object (are the N/S 'smearings' a reflection of pulse width and not debris dimension ?). Noting its a jpeg, any need to be careful about compression artefacts ?

Grity: never heard of Lilienthal: impressively ahead of his time !

The image of fuselage section looks like it supports a limited front/aft impact load, and if people/seats and attachments are intact, that would also indicate realtively limited impact loadings.

Is the linear debris trace trending roughly north from the centre taken to be the remains of a late breaking stb wing ?

It would appear to me to be completely fruitless to attempt ANY 'analysis' of heading at impact from the lie of any wreckage. The remains of the aircraft will 'fly' just as well in water as in air, and since most seem to be postulating some sort of rotating motion at impact there is no reason why this should not continue during the 4000m drop resulting in ANY sort of orientation. Until FDR data is recovered, all the wreckage tells us at the moment is roughly where it hit the surface.

Fair comment, stb was perhaps pushing it. First thought was it was a debris trail as if both engines have been pictured, where is the other wing.

With respect to other recent musings, re-reading the BEA report reminded me that a concurrent flight had to carefully adjust the weather radar to get the full picture and that no big issue is made of the difference between ACARS 'WRN' and 'WRG' : it doesn't reference interconnection failure for the latter, just a mistrust.

'Morning BOAC,

Quote:
"It would appear to me to be completely fruitless to attempt ANY 'analysis' of heading at impact from the lie of any wreckage."
I'm not sure you or I know enough to make that assumption. Let's keep on interrogating the experts, just in case. As auraflyer (while admitting to be a layperson) says: "The pieces which are both densest and offer least resistance, descend fastest and move least with current. Pieces that are lightest and with the most resistance take the longest to fall and are pushed furthest by the current. Other things fall in between."

Comparing the pattern of returns in the east and west, I have a hunch that swordfish41 and others may resolve the 250/070 dilemma at a glance. However, as I asked earlier, is the orientation of this debris field necessarily indicative of the heading ar sea-level impact?

Quote:
"The remains of the aircraft will 'fly' just as well in water as in air, and since most seem to be postulating some sort of rotating motion at impact there is no reason why this should not continue during the 4000m drop resulting in ANY sort of orientation."
Cannot agree with any of that.

Quote:
"...all the wreckage tells us at the moment is roughly where it hit the surface."
It depends what you mean by roughly. For a start, the engines - presumably both detached at sea-level - would have gone pretty-much straight to the bottom. As mm43 commented some time ago, their position relative to this debris field is of the essence.

Maybe the BEA will divulge the Lat/Long of this debris field eventually. But I, for one, am not holding my breath. Pity: I can't see any third party being able and inclined to interfere.

Sonar interpretation
 

Sorry to get back a bit late into this. I did say "hazard a guess," and it could be the object is the tail section, although I assumed that the swathe width was 700m, and that the altitude of the AUV was therefore much higher than 60m, suggesting that the object was also higher above the seafloor than auv-ee suggests. If it is the tail I expected to see an equaly large object at the western end of the debris trail, and I was influenced in this by mm43's graphic of the likely break up of the fuselage at impact.I disagreed with mm43's view that the the debris would be fairly well confined, and I was wrong about that too. However the two Remus vehicles will have completed the photo mosaic now and BEA will know just where everything is. I've followed all the threads from this incident from day one, and incredibly fascinating its been. Throughout I have had at the back of my mind a story told me by the Chief Lab Technician on Odin Finder. He was engaged to search for an RAF Jaguar that crashed in the Adriatic. The pilot had ejected quite late, and confidence was high that the position of the aircraft on the sea floor was known. After no trace was found he fielded the usual questions, "Is your equipment working properly?" "Have your technicians missed anything?" with his own, "where do you want the ship to go now, north, south, east or west?" My granny told me when I lost something that it would be in the last place that I looked.

With respect to orientation, on reflection agree with BOAC's chastisement. However, the assymetry N/S, with apparent distribution linearly to the north not reflected to the south may reflect something. But perhaps if you stare long enough into the darkness you will eventually see a ghost. Hoped for a clear debris trail.

Ghosts
 

Mr Optimistic has hit the nail on the head. The great danger is to see what you want to see.

swordfish41,

Further to previous posts by me and others, is it possible to say if the 070/250 orientation of the debris field has been created purely by the prevailing currents during the descent, or is it in any way indicative of the aircraft's heading at sea-level impact?

Not 'heading' in my book. By the way, I'm puzzled why do you not view water as a liquid? "Cannot agree with any of that."

Debris orientation/position at 4000m will be governed only by the currents and the 'flying'/break off motion of individual separate parts and there will be a ZERO effect of any motion vector at impact. All bits will come to a pretty abrupt halt at immersion. Water is hard at that speed. As said, separated engines will go pretty much straight down. The fin/seat cushions etc float. Plot all points between?

BOAC,

I cannot DISagree with any of that! But it is completely at odds with the sentence I criticised in your previous post:
"The remains of the aircraft will 'fly' just as well in water as in air, and since most seem to be postulating some sort of rotating motion at impact there is no reason why this should not continue during the 4000m drop resulting in ANY sort of orientation."

Re your latest: would you prefer "track"? Not sure the difference is significant in this case.

Heading/track no matter. Neither apply. 'Track' becomes meaningless with no overall forward speed. Break off part of a wing and see what happens to 'heading' as the wreckage sinks. You can, in my opinion, deduce nothing.

My post is actually COMPLETELY in agreement with my previous. Hydrodynamics work pretty much as aerodynamics at low speeds. A wing will produce a force of some sort. An attached tailplane will produce a pitching motion as the wreckage sinks.

Next time you fill a deep bath, generate a gentle 'current' by moving your hand around in the water. Then throw in a small strip of metal and see if you can deduce from where and how it sits on the bottom which way it was travelling when it hit the water - or how fast, for that matter. Now imagine that bath thousands of times deeper.

Shadoko, you may be right that the bodies recovered were those not seated. However, on the old thread a fellow involved in recoveries from water suggest that dead bodies seem to work their way out even from tight seat belts when you have water to work with. They even work their way out of their clothes much of the time.

Given how small the pattern is I suspect those recovered may have been in seats such as those recovered. But they did not float to the surface from a seat still attached to a large piece of fuselage.

I find it interesting so many were not sitting and not strapped in. Were they holding a dance in the aisles for entertainment? It's another of the surprises associated with this accident.