AF447
Join Date: Jun 2009
Location: Paris
Posts: 691
Likes: 0
Received 0 Likes
on
0 Posts
It is incredible the amount of imagination caused in people minds by this vertical stabilizer wreckage recovered. One wonder if they are realizing that if the hydraulic system was cripled for whatever reason, and it will be in any case of structural damage implying the rudder, the maintenance computer won't simply address a four minute string of unreliable speed related problems...
On the other hand, when two different materials are assembled together, each one with singularly different resistance and deformation behavior, it is not very hard to understand that a separation will very likely result once a critical level of acceleration is applied.
On the other hand, when two different materials are assembled together, each one with singularly different resistance and deformation behavior, it is not very hard to understand that a separation will very likely result once a critical level of acceleration is applied.
Recovered wreckage
I have tried to search but with 155 pages....
What physical evidence there is, compared to other 'flat spin accidents' pointed out here, seems to show a surprising lack of 'component' as opposed to 'structural' damage (eg the VS separated at the root but little other crumpling, the weak galley structure with right angle joints still intact and pristine trays, the section of wall/bulkhead with jump seats, the control surface with a longtitudinal tear), also no clear (in these views) witness marks of secondary impacts with other translating objects.
Has anyone posted a view of the a/c showing the likely position of the recovered parts on the airframe and were all these recovered in the same general patch of ocean ?
Also worth bearing in mind that the sea wasn't flat: initial impact forces may have been concentrated on certain parts of the structure.
If this has been done over a thousand times here I know what to expect
What physical evidence there is, compared to other 'flat spin accidents' pointed out here, seems to show a surprising lack of 'component' as opposed to 'structural' damage (eg the VS separated at the root but little other crumpling, the weak galley structure with right angle joints still intact and pristine trays, the section of wall/bulkhead with jump seats, the control surface with a longtitudinal tear), also no clear (in these views) witness marks of secondary impacts with other translating objects.
Has anyone posted a view of the a/c showing the likely position of the recovered parts on the airframe and were all these recovered in the same general patch of ocean ?
Also worth bearing in mind that the sea wasn't flat: initial impact forces may have been concentrated on certain parts of the structure.
If this has been done over a thousand times here I know what to expect
Last edited by Mr Optimistic; 5th Jul 2009 at 13:01. Reason: addition
Join Date: Aug 2006
Location: London
Posts: 51
Likes: 0
Received 0 Likes
on
0 Posts
IMHO, this means also, that due to High and Low Speed Stabilization (where the aircraft gives inputs to the flight controls to prevent stalling it [and therefore, spin it]) and also to the fact that AoA data is provided to BUSS - Back Up Speed tape, via the IR part of the ADIRU, in order to enter a stall the aircraft would have to have changed to Direct Law.
Join Date: Jun 2009
Location: Somewhere out there
Age: 39
Posts: 65
Likes: 0
Received 0 Likes
on
0 Posts
AOA - IR
and also to the fact that AoA data is provided to BUSS - Back Up Speed tape, via the IR part of the ADIRU, in order to enter a stall the aircraft would have to have changed to Direct Law.
P.S. - Who knows whether they (AF447) went back to normal law or not...
EDITING: The AoA is normally provided by the ADRs - The "BUSS" provides it through IR as a backup. So the doubt remains as to the role of a defective IR in the "standard" AoA output
Last edited by augustusjeremy; 5th Jul 2009 at 13:50.
Join Date: Dec 2006
Location: Italy
Age: 45
Posts: 46
Likes: 0
Received 0 Likes
on
0 Posts
Of course not, because airliners don't spin.
Marco
Join Date: Nov 2006
Location: SoCalif
Posts: 896
Likes: 0
Received 0 Likes
on
0 Posts
It is a spin when one wing is stalled more than the other. The rotation results from the less-stalled wing providing more lift than the other.
A stall occurs when both wings stop flying about equally. Recovery occurs if the cg is ahead of the C/L, center of lift, as the nose will drop, and flying speed is regained.
If the cg is behind the C/L of stalled wings, the nose will not drop, and the plane will not regain flying speed. This is a flat spin. It does not require rotation. As above, for rotation there has to be a partly unstalled wing. Per Wiki, the flat spin angle of attack will be between 65 and 90 degrees, which means there could be some forward speed.
Modern airliners are intentionally flown at aft cg near the safe limit. A load shift will put the cg behind the C/L, and this has happened countless times, usually with cargo.
GB
A stall occurs when both wings stop flying about equally. Recovery occurs if the cg is ahead of the C/L, center of lift, as the nose will drop, and flying speed is regained.
If the cg is behind the C/L of stalled wings, the nose will not drop, and the plane will not regain flying speed. This is a flat spin. It does not require rotation. As above, for rotation there has to be a partly unstalled wing. Per Wiki, the flat spin angle of attack will be between 65 and 90 degrees, which means there could be some forward speed.
Modern airliners are intentionally flown at aft cg near the safe limit. A load shift will put the cg behind the C/L, and this has happened countless times, usually with cargo.
GB
Join Date: Apr 2002
Location: FUBAR
Posts: 3,348
Likes: 0
Received 0 Likes
on
0 Posts
Have to agree with petermcleland in the post above, it is absolutely imperative the Brazilians furnish the BEA with the findings of the post mortems or that a further post mortem takes place following the release of the bodies.
There has been enough historical data I am sure to correlate the type of injuries and cause of death with a particular scenario, or at least to help in discounting other theories, which could be invaluable until more concrete data (hopefully) becomes available from further wreckage recovered, or with some chance the recorders.
Any idea why the Brazilian authorities wouldn't make this info available to the BEA as soon as the work had been completed ? , strange indeed
There has been enough historical data I am sure to correlate the type of injuries and cause of death with a particular scenario, or at least to help in discounting other theories, which could be invaluable until more concrete data (hopefully) becomes available from further wreckage recovered, or with some chance the recorders.
Any idea why the Brazilian authorities wouldn't make this info available to the BEA as soon as the work had been completed ? , strange indeed
Lithium battery life
Originally posted by Takata
This is because of the properties of the lithium battery, I think. The nominal voltage stays the same for the specified 30 days and then some, but once the voltage starts to drop, the drop is rather dramatic and therefore the beacon stops sending any acoustic output at some point, rather sooner than later.
Extended battery life would not be too difficult to achieve. Another 200 grams of additional weight perhaps. Peanuts, really.
The manufacturer opinion is that the pinger signal is expected to fade very quickly after operating for 30 days
Extended battery life would not be too difficult to achieve. Another 200 grams of additional weight perhaps. Peanuts, really.
Join Date: Oct 2006
Location: Gone Flying...
Age: 63
Posts: 270
Likes: 0
Received 0 Likes
on
0 Posts
BUSS - Back Up Speed SCALE 
Graybeard:
What law was controlling the flight test A330 when it stalled and crashed?
Join Date: Jun 2009
Location: SPAIN
Age: 65
Posts: 50
Likes: 0
Received 0 Likes
on
0 Posts
Battery life
Hi,
http://www.pprune.org/tech-log/37905...tery-life.html
Bye.
Extended battery life would not be too difficult to achieve. Another 200 grams of additional weight perhaps. Peanuts, really.
Bye.
Join Date: Jun 2009
Location: Belfast
Age: 46
Posts: 21
Likes: 0
Received 0 Likes
on
0 Posts
Humble Pie
My apologies for wading in to the discussion a while ago, highlighting similarities in the condition of bodies with previous mid-air accidents.
I've been put in my place by the recent report, and realised that whilst I (like the rest of you) am very concerned and interested in what happened that night, I am not close to the investigation, I am not privvy to all the available information, and as such I really shouldn't have 'waded in' with yet another peice of conjecture.
I was impressed with the honesty and detail of the prelim report, and I now have full confidence that the experts are doing their very best. Let's leave them to do their jobs.
In the mean-time I think we're just going to need some patience.
I've been put in my place by the recent report, and realised that whilst I (like the rest of you) am very concerned and interested in what happened that night, I am not close to the investigation, I am not privvy to all the available information, and as such I really shouldn't have 'waded in' with yet another peice of conjecture.
I was impressed with the honesty and detail of the prelim report, and I now have full confidence that the experts are doing their very best. Let's leave them to do their jobs.
In the mean-time I think we're just going to need some patience.
Join Date: Jun 2009
Location: Florida
Posts: 17
Likes: 0
Received 0 Likes
on
0 Posts
Phil Gollin - FDR/CVR search & recovery
I don't know about the status of the search, or about the 'true' battery life of the devices. Other people have commented on these issues. I used to work in the hydrographic survey industry, so I can comment on finding AF447 in the event that the pingers aren't detected.
A quick look on Google Earth at the seafloor in the region of the accident shows that it's pretty rugged. This means that finding the wreckage using the usual methods (sidescan sonar) might be tricky - boulders, cliffs, etc all give shadows which can look like wreckage, or the wreckage return can be hidden in those shadows. If you spent enough money doing surveys in all directions (so that the shadows are cast in different directions in each survey), and were scrupulous about investigating all potential contacts and databasing them ... eventually you would be 99% certain to find it. How long would it take & how much money? Could be 'a long time' and 'a lot'. I don't know the size of the search area, so I couldn't begin to estimate, but certainly tens of millions of dollars, and hundreds if it goes on for very long. You'd need deepwater ROVs to video each contact, and the day rates for the vessels which deploy those are very high (like $250k per day, IIRC).
If the wreckage is largely in one place, which may be the implication of the BEA report, it would seem sensible to continue searching. If it had fragmented in flight, chances of finding specific individual pieces (e.g. FDR, CVR) would be pretty poor. (I mean: you could spend $1bn and not find the CVR & FDR; so you would have to make a judgement up front as to whether it's worth doing the search, based on whether you think the debris field will be localized or scattered over tens of square miles).
I guess some people would have a view on whether it's in the interests of the French gov't to try very hard to find the wreckage. (I don't believe that sort of thing for a second, myself ... maybe I'm just naive).
A quick look on Google Earth at the seafloor in the region of the accident shows that it's pretty rugged. This means that finding the wreckage using the usual methods (sidescan sonar) might be tricky - boulders, cliffs, etc all give shadows which can look like wreckage, or the wreckage return can be hidden in those shadows. If you spent enough money doing surveys in all directions (so that the shadows are cast in different directions in each survey), and were scrupulous about investigating all potential contacts and databasing them ... eventually you would be 99% certain to find it. How long would it take & how much money? Could be 'a long time' and 'a lot'. I don't know the size of the search area, so I couldn't begin to estimate, but certainly tens of millions of dollars, and hundreds if it goes on for very long. You'd need deepwater ROVs to video each contact, and the day rates for the vessels which deploy those are very high (like $250k per day, IIRC).
If the wreckage is largely in one place, which may be the implication of the BEA report, it would seem sensible to continue searching. If it had fragmented in flight, chances of finding specific individual pieces (e.g. FDR, CVR) would be pretty poor. (I mean: you could spend $1bn and not find the CVR & FDR; so you would have to make a judgement up front as to whether it's worth doing the search, based on whether you think the debris field will be localized or scattered over tens of square miles).
I guess some people would have a view on whether it's in the interests of the French gov't to try very hard to find the wreckage. (I don't believe that sort of thing for a second, myself ... maybe I'm just naive).
Join Date: Sep 2000
Location: England
Posts: 303
Likes: 0
Received 0 Likes
on
0 Posts
The Role of Auto-trim?
The one very real way in which you can complicate your recovery from a loss of control (spin or whatever) is to have the autopilot surreptitiously auto-trim in as much back-trim as possible. However the pilot doesn't need to organize that, the autopilot will do that all by itself, quite unnoticed by the pilot - as the speed erodes. That pitfall of an autopilot disguised contrary control force is why medium size turboprop airliners are required to be handflown in heavy icing conditions. The phenomenon of autopilot kickout immediately followed by autorotation (i.e. a spin) has been the undoing of many an inattentive crew - after a short period of ice accumulation. We have the recent examples of the DHC8-400 in the US and the Turkish 737 at Amsterdam to emphasize the stealthy suckertrap of auto-trim.
But what was happening to the THS (trimmable horizontal stabilizer) and auto-trim aboard AF447 when the pitot malfunction occurred? Could that pitot icing indication glitch have emulated the icing performance loss scenario? Can the pitot's induced false speed reading predicate the auto-trim's response? Can the auto-trim then be well out of whack nose-up and yet the autopilot still soak up the "stick"-load in order to maintain the stipulated altitude?
If this was in fact (or could've been) the case, what would have happened when autopilot, auto-thrust and (presumably) auto-trim dropped out? How difficult (or improbable) would recovery be if the autopilot kicked out due to it holding greater than a kick-out threshold force of forward elevator?
I imagine a pitch-up into a stall and a wing-drop flick into autorotation would likely be an instant development. Would the pilot, now in manual control and Alternate Law, have any clue as to what was happening? - such that he could correct the out-of-trim configuration and regain control?
Or is it more likely that the combination of unaccustomed manual flight at high altitude, sudden IMC, Alternate Law and roll/pitch/yaw could mask the
back-trim and leave the pilot totally flummoxed?... You only need scant seconds of very nose-low to exceed VNe. And of course we're disregarding here what could have happened to one or both engines thrust during that pitch-up - which may have provided the asymmetry for spin entry.
I've got first hand experience of what happens to a JP5A (Jet Provost) pilot subjected to loads of elevator back-trim when handed over to for spin recovery. They were always non-plussed. The learning experience exceeded the syllabus for spinning but similar to inspin and outspin aileron's effects, it was a worthwhile exercise to file away in instant recall.
Is this statement still operative once the speed input to the ADIRS is wholly corrupted? The aerodynamics of a pitch-up at high altitude would trump that law - methinks.
More food for thought. Think centrifugal force in a large aircraft spin. Even if it wasn't a flat spin to begin with, anybody/anything loose in the cabin will migrate towards the rear, pushing the CofG even further AFT. Even with baffles, this might also be the case for some of the fuel. Any thrust left on the engines will also tend to flatten the spin. Any heavily back-trimmed aircraft will tend to want to remain in a flat spin, particularly as the thrust increases with height loss (for any given set throttle lever position).
.
But what was happening to the THS (trimmable horizontal stabilizer) and auto-trim aboard AF447 when the pitot malfunction occurred? Could that pitot icing indication glitch have emulated the icing performance loss scenario? Can the pitot's induced false speed reading predicate the auto-trim's response? Can the auto-trim then be well out of whack nose-up and yet the autopilot still soak up the "stick"-load in order to maintain the stipulated altitude?
If this was in fact (or could've been) the case, what would have happened when autopilot, auto-thrust and (presumably) auto-trim dropped out? How difficult (or improbable) would recovery be if the autopilot kicked out due to it holding greater than a kick-out threshold force of forward elevator?
I imagine a pitch-up into a stall and a wing-drop flick into autorotation would likely be an instant development. Would the pilot, now in manual control and Alternate Law, have any clue as to what was happening? - such that he could correct the out-of-trim configuration and regain control?
Or is it more likely that the combination of unaccustomed manual flight at high altitude, sudden IMC, Alternate Law and roll/pitch/yaw could mask the
back-trim and leave the pilot totally flummoxed?... You only need scant seconds of very nose-low to exceed VNe. And of course we're disregarding here what could have happened to one or both engines thrust during that pitch-up - which may have provided the asymmetry for spin entry.
I've got first hand experience of what happens to a JP5A (Jet Provost) pilot subjected to loads of elevator back-trim when handed over to for spin recovery. They were always non-plussed. The learning experience exceeded the syllabus for spinning but similar to inspin and outspin aileron's effects, it was a worthwhile exercise to file away in instant recall.
in order to enter a stall, the aircraft would had to have changed to Direct Law.
If the cg is behind the C/L of stalled wings, the nose will not drop, and the plane will not regain flying speed. This is a flat spin. It does not require rotation. As above, for rotation there has to be a partly unstalled wing. Per Wiki, the flat spin angle of attack will be between 65 and 90
degrees, which means there could be some forward speed.
Modern airliners are intentionally flown at aft cg near the safe limit. A load shift will put the cg behind the C/L, and this has happened countless
times, usually with cargo.
degrees, which means there could be some forward speed.
Modern airliners are intentionally flown at aft cg near the safe limit. A load shift will put the cg behind the C/L, and this has happened countless
times, usually with cargo.
.
Last edited by TheShadow; 5th Jul 2009 at 15:25.
Join Date: Jun 2009
Location: Somewhere out there
Age: 39
Posts: 65
Likes: 0
Received 0 Likes
on
0 Posts
calcs ?
100mph horizontal does surprise me. I would be interested in seeing your math, pm it to me if you don't mind.
radius of spin circle: 29m
circumference: 2piR (pi =~ 3.14etc) =~ 2 * 3.14 * 29 = 182.12 m
182.12/4 (a 'complete' spin each 4s) = 45.53m/s (the tail is running along the path of the circumference)
45.53m/s = 45.53m * 3600/h =~ 164km/h
aprox. 100mph
Join Date: Apr 2009
Location: Petaluma
Posts: 330
Likes: 0
Received 0 Likes
on
0 Posts
augustusjeremy
4 seconds per rotation = ~175'/sec.
175 feet/sec. = ~10,500'/minute = ~ 115 mph.
115 mph with an axis w/o horizontal velocity.
With horizontal velocity, add for lead, subtract for lag.
The wing tip or tail could have a velocity through air approaching 200mph.
4 seconds per rotation = ~175'/sec.
175 feet/sec. = ~10,500'/minute = ~ 115 mph.
115 mph with an axis w/o horizontal velocity.
With horizontal velocity, add for lead, subtract for lag.
The wing tip or tail could have a velocity through air approaching 200mph.
Join Date: Jun 2009
Location: Somewhere out there
Age: 39
Posts: 65
Likes: 0
Received 0 Likes
on
0 Posts
momentum
augustusjeremy
4 seconds per rotation = ~175'/sec.
175 feet/sec. = ~10,500'/minute = ~ 115 mph.
115 mph with an axis w/o horizontal velocity.
With horizontal velocity, add for lead, subtract for lag.
The wing tip or tail could have a velocity through air approaching 200mph.
4 seconds per rotation = ~175'/sec.
175 feet/sec. = ~10,500'/minute = ~ 115 mph.
115 mph with an axis w/o horizontal velocity.
With horizontal velocity, add for lead, subtract for lag.
The wing tip or tail could have a velocity through air approaching 200mph.
I am not an aerodynamicist but maybe the momentum exerted by the vertical stabilizer in this context in such a large (long) aircraft explains why they're harder to put into a flat spin.
It would depend on the center of rotation/gravity/mass, I believe
Join Date: Apr 2009
Location: Petaluma
Posts: 330
Likes: 0
Received 0 Likes
on
0 Posts
Given the pic of the Fuji B707 and B-70, (not that they are representative), rotation in a fully developed flat spin would seem to be almost nonexistent. Any rotation would add asymmetric lift (especially to a swept wing design) and tend to revert the a/c to a conventional spin.
BEA's statement re: flat and nearly vertical impact is sustained by the reasoning that dictates virtually nil or nil rotation.
BEA's statement re: flat and nearly vertical impact is sustained by the reasoning that dictates virtually nil or nil rotation.