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

AF447

Old 10th Aug 2009, 02:25
  #4181 (permalink)  
 
Join Date: Nov 2005
Location: Harvest, Alabama
Posts: 109
Likes: 0
Received 0 Likes on 0 Posts
I once crossed 11000' over the top of a line just west of Chicago. I was young and dumb, and a senior guy was in the left seat. We started at FL370. Big bump, and we started going up. Throttles closed, A/P off, nose down as far as we dared, MMO approaching, boards up, up and away we went.

Topped out at FL440 in a G2, idle power, boards up (Ceiling FL450).

Chicago center had no one above us (it was ISA+10) as I remember (Thank you Lord). They shut anyone from following us after that. The guy in front of us busted his service ceiling (an MD80 that started at FL330).

We drifted back down to FL 370 with the boards up, throttles closed, and tried to stabilize the cabin altitude.

After that, I had a personal rule about crossing rapidly building lines.

Don't.

The 47000' rumor for an A330? As clean as that airframe is, above SVC ceiling, at night, in or above wx, no horizon (visual or artificial), possibly no A/S display?

No way you'd walk away from that one. But then we already knew that.
singpilot is offline  
Old 10th Aug 2009, 06:58
  #4182 (permalink)  
 
Join Date: Apr 2002
Location: Surrey, UK ;
Age: 71
Posts: 1,150
Received 7 Likes on 7 Posts
47,000 ???? Where's that come from

I am just dumbfounded - As there is nothing I have seen or read to even start a rumour of AF447 getting to FL 470. We have no survivors, no FDR or CVR. no radio calls, no radar coverage and Tim Vasquez's weather evaluation all conspiring against it.

Lets keep speculation within reasonable limits or state some logic behind the rumour.
Dave Gittins is offline  
Old 10th Aug 2009, 07:03
  #4183 (permalink)  
PJ2
 
Join Date: Mar 2003
Location: BC
Age: 76
Posts: 2,483
Received 0 Likes on 0 Posts
singpilot;

It's the one scenario that could explain the ACARS Cabin Alt message. That said, all here would certainly require more data and it just hasn't appeared and hasn't been posited as "something new" from the ACARS messages. Other than being mentioned in the list and the limits explained, it's not dealt with in the section explaining the messages.
PJ2 is offline  
Old 10th Aug 2009, 09:14
  #4184 (permalink)  
 
Join Date: Dec 2007
Location: Itinerant
Posts: 825
Received 45 Likes on 10 Posts
DG -- While I agree that there is nothing to substantiate a specific altitude as high as FL470, there is considerable relevant logic that supports a hypothesis of a rapid climb significantly past, say, FL 410.

I have seen nothing that would rule out a thermally driven rapid climb in the given circumstances, whereas I have certainly seen many posts on this thread suggesting scenarios that are either highly implausible, or even ruled out.

As PJ2 suggests, such a scenario is one possible explanation for one of the as yet unexplained ACARS messages.

FWIW, I am one of those who is still somewhat optimistic with regard to the data recorders being located. Soon I hope.
grizzled is offline  
Old 10th Aug 2009, 10:36
  #4185 (permalink)  
 
Join Date: Jan 2008
Location: Herts, UK
Posts: 748
Likes: 0
Received 0 Likes on 0 Posts
Could it be, not the ACARS messages, but the method of transmission & reception that offers such data...
I have no knowledge of this technology and from what I've read previously and reasoning that 2 miles in altitude would require timing to absurd limits, triangulation too, can only ask that experts here knock my suggstion back quickly and categorically!
HarryMann is offline  
Old 10th Aug 2009, 11:12
  #4186 (permalink)  
 
Join Date: Apr 2002
Location: Surrey, UK ;
Age: 71
Posts: 1,150
Received 7 Likes on 7 Posts
Snoop

grizzled I am not immune to any well reasoned hypotheses but having just read Tim Vasquez's analysis again, he draws the conclusion that whilst nothing can be ruled out, there is no evidence of significant or extraordinary activity in the ITCZ that night or any unusual updrafts. His diagram at fig 13 shows the 447 just through an area of suspected updrafts but he also notes that updraft strength is generally lower in oceanic Cbs than overland Cbs.

That said, if I experienced a sudden and dramatic updraft I would likely put the left wing down and try to make a descending 180 .. which perhaps tends towards the "turnback" theory popular a few thousand posts back.

As you say, roll on the FDR & CVR being found ... it's surely only a matter of time.
Dave Gittins is offline  
Old 10th Aug 2009, 11:35
  #4187 (permalink)  
 
Join Date: Jan 2008
Location: Herts, UK
Posts: 748
Likes: 0
Received 0 Likes on 0 Posts
...he draws the conclusion that whilst nothing can be ruled out, there is no evidence of significant or extraordinary activity in the ITCZ that night or any unusual updrafts.
Maybe not unusual from a met aspect, but is it common to fly through the centre of such updrafts, which he also admits I believe, could have been topping out at 51,000 ft, having burst through the surrounding tropopause ...

which could have happened that dark and stormy night?
HarryMann is offline  
Old 10th Aug 2009, 11:45
  #4188 (permalink)  
 
Join Date: Jan 2008
Location: Herts, UK
Posts: 748
Likes: 0
Received 0 Likes on 0 Posts
IF...

Another outside odds 'fit' might be that having to fly pitch and power once A/S seemed unreliable, if they followed AFM and did so, and if they realised their climb-rate, pulling the thrust back would have been 'difficult to justify', at least intially.
HarryMann is offline  
Old 10th Aug 2009, 13:05
  #4189 (permalink)  
 
Join Date: Jun 2009
Location: Oxford, England
Posts: 297
Received 0 Likes on 0 Posts
I think I understand what you mean and I think we understand that there may be as much a philosophical meaning as a semantic one here. No software is "aware". My very simple understanding of software is, while software can mimic learning, software doesn't "learn" in the meaning of the term we usually understand.
Semantics… No computer has real ‘intelligence’, but systems can be designed to be adaptive or more or less ‘aware’ as required, for a particular application. The problem is that the more complex the algorithm, the more likely there are to be corner cases that the designers didn’t consider and there are many complex interacting variables involved in the control of an ac. Thus, the kiss principle applies and is especially relevant in safety critical applications since the simpler the system, the easier it is to test and show that it is demonstrably correct and deterministic.


For a number of reasons, I am not sure that "fuzzy logic" solutions are suitable to airline work.
A sort of half way house between a completely dumb control system and ai. At a basic level, it takes the weighted values of many inputs and perhaps time to compute an output value or decision.

An example of where such an approach might be usefull in avionics is the pitot failure situation. Assume level flight for some period, constant as and with ground speed from gps and perhaps doppler. A computer keeps track of these sensors, their rates and their relationship over a several minute period, as well as other variables of interest such as aoa, vertical speed, attitude etc. The historical set and running averages can be used to predict the next set of values to a fair degree of accuracy over a short timescale. If any single value falls outside a defined window, or exceeds rate limits at next sample time, the system may resample to filter noise before rejecting the sensor. It then makes a best effort estimate from remaining sensors and presents that to the user, together with alarms for the suspected failing sub system and degraded accuracy of data. Such an approach becomes even more relevant where the outputs from one subsystem become dependent inputs for others. ie: systems can and should be designed to prevent domino effect failure. The key thing is that, whatever the overall system design, it should degrade gracefully and produce unambiguous data at all times. It fails completely if it is unable to do this, or gives up in such a way as to present the controlled entity to the user in an unknown state...

Chris

Last edited by syseng68k; 10th Aug 2009 at 14:48.
syseng68k is offline  
Old 10th Aug 2009, 14:54
  #4190 (permalink)  
 
Join Date: Apr 2009
Location: A place in the sun
Age: 82
Posts: 1,250
Received 44 Likes on 17 Posts
I retired from active flying nearly 15 years ago, so I must not fall into the trap of thinking that more ‘stick and rudder’ flying is the only panacea for some of the problems that are manifesting themselves today. The last aircraft I flew was the B747-100/200 series!

Having been out of the loop for the last few weeks I have missed much of the discussion in PPRUNE about AF447. Now that I have had the chance to read back through some of the comments in this thread it seems to me that we are in danger of straying from the essentials.

First, we still don’t know what happened. There is some evidence that the Thales pitot tubes may have been at fault. Equally, I found Tim Vasquez’s analysis of the weather very informative (see www.weathergraphics.com/tim/af447/ ). But it is still all speculation.

Second, we must accept that however good aviation systems have become there are phenomena out there that exceed anything we can reasonably expect to survive. I have recently been sailing along the south coast of Ireland and have been reminded of the so-called ‘freak waves’ which can in some cases reach heights of 30 mtrs. It is worth remembering that, world-wide, over the last two decades, more than 200 super-carriers - cargo ships over 200m long - have been lost at sea. Eyewitness reports suggest many were sunk by high and violent walls of water that rose up out of calm seas. In the air, thunderstorms have the power to wreak similar havoc and must be treated with the greatest of respect.

Third, while reading this thread, the emphasis has shifted from what little is known about AF447 to a more general and very valid discussion on handling skills and automation. Harry Mann and PJ2 discussed stalling (6 Aug). My view is that current pilots do not know enough about the stall characteristics of their aircraft simply because this is treated at such a superficial during training. D.P. Davies (Handling the Big Jets, 3rd edition, page 110) suggests that aircraft achieve stall warning speeds on somewhere between 1 in 100 and 1 in 1000 flights; and stall speed on around 1 in 100,000 flights. When he wrote this book he also compared this with engine failure rates at that time of approx 1 in every 1000 flights en-route and 1 in 100,000 on take-off near V1. Since then, engines have become much more reliable but we still seem to have accidents caused by stalling. I would argue that, in modern training, more emphasis should be placed on handing stalls. I used to do C. of A. test flights on VC10s, B707s and B747s and was surprised at how many line pilots were so concerned about the safety of such manoeuvres. Clearly this was because of their comparative lack of exposure. The controlled conditions of a test flight are one thing, with few surprises. At FL350 on a dark night in turbulence it would be another matter, but I am sure more adequate simulator training would improve skills and increase confidence.

Fourth, with so much of the operation now being done using automation, it is clear that situational awareness and manual handling skills are declining. Phantom Driver (6 Aug) quite rightly says that he does not want to have a wild ride while some guy up front tries to polish up his handling skills. PJ2 (also on 6 Aug) highlights how little manual handling is done these days on the route and that next to NO time is spent on hand-flying on the simulator. He has also written at length (8 Aug) on the positive effect that modern automation has had on flight safety – if we went back to the old methods I am convinced that the accident rate would increase. Therefore, the issue is to decide what training is relevant to the modern situation. I think that PJ2, Harry Mann and Phantom Driver would all agree that what is needed is a thorough analysis of exactly what needs to be trained and then to revise conversion and recurrent training schedules so that they reflect more closely the current risks rather than those of yesteryear. I also think that more hand-flying in the simulator, with exercises specifically designed to increase confidence in taking over manually after unusual problems, would be of great benefit to increase confidence and manual skills. Perhaps, exercises of this sort would have helped the Turkish crew at Amsterdam.

Fifth, one final thought – this thread and many others following an accident tend to wear a ‘hair shirt’. Accidents always hit the headlines, and then we rake over the coals at great length. However, the quiet, competent saving of an extremely hazardous situation is hardly noticed – unless it happens under the gaze of the press on the Hudson River! Perhaps we should spend a little more time celebrating the remarkable things that are achieved by our colleagues every year.
Bergerie1 is online now  
Old 10th Aug 2009, 16:51
  #4191 (permalink)  
 
Join Date: Jan 2001
Location: France
Posts: 168
Likes: 0
Received 0 Likes on 0 Posts
EASA Proposed Airworthiness Directive for Pitot tube replacement on A330/340

10/08/09 Number 09-099
http://ad.easa.europa.eu/blob/easa_p...f/PAD_09-099_1
Squawk_ident is offline  
Old 10th Aug 2009, 17:34
  #4192 (permalink)  
 
Join Date: Jan 2006
Location: Singapore
Posts: 320
Received 0 Likes on 0 Posts
Bergerie 1:

Hopefully some Heads of Training will take in these points. The sad fact is, too many of us old timers (yours truly included) and the younger ones weaned on automation, have forgotten how to "fly"!

I always recall a DC10 accident some years ago at (Boston?)--short(ish) runway, wet snow; auto thrust system had a known history of faults. On this occasion, it held Vref+20 all the way down the approach. This was noted and commented upon by the crew, but nothing was done to rectify the situation. A/C landed hot and aquaplaned off the end in to the river.

The accident report made note of the fact that the Captain had not flown with A/T out for the past 5 years and so was reluctant to disconnect and fly manual throttles. Similar ingrained habits seen in the Qantas 744 incident at Bangkok; idle reverse was (apparently) the current SOP, resulting in another overrun on a wet runway following an aborted G/A (Captain over riding F/O's decision?).

David Beaty had some interesting things to say about all this in his "Human Factors in Air Accidents". In stressful times we all, of course, tend to revert to the habit which is most comfortable, as tunnel vision sets in. A pity really,because the lessons have always been there. As some old sage (Samuel Johnson, if memory serves correctly) wrote-"Man has oft more need to be reminded than informed".

In the Air Force (love ém or hate ém!}, the instructors used to delight in drilling into us -"You Fight as you Train"; seems like the same lessons could well be carried across into our Airline ops. We really do owe it to the customers.
Phantom Driver is offline  
Old 10th Aug 2009, 18:09
  #4193 (permalink)  
 
Join Date: Jun 2009
Location: in a plasma cocoon
Age: 53
Posts: 244
Likes: 0
Received 0 Likes on 0 Posts
Originally Posted by grizzled
I have seen nothing that would rule out a thermally driven rapid climb in the given circumstances, whereas I have certainly seen many posts on this thread suggesting scenarios that are either highly implausible, or even ruled out.
Hi there !
Please, could you be more specific about the either implausible or ruled out scenarii ? On my own, I have seen nothing that would acertain this hypothesis of a rapid climb at FL470 (why 470 ? no overshoot in the thermal imagery, no updraft detected by the AMDAR flight 25 mn before on the same track but at FL325 - wind 3.6m/s@319° , conditions can change rapidly though). Does this scenario of a rapid climb even require a freezing event of the Pitots ? It would be a coincidence ? (Pitot freezing event & rapid/uncontrolable climb)
Where would be the problem with a possible unfortunate reaction of the automation and the pilots in a context of unreliable airspeeds, misleading signals in the cockpit and in the procedures, high workload and false stall alarms, that would have degraded the aerodynamic margin at high altitude and rendered the plane highly vulnerable to turbulence even moderated ? (loss of control in a Learjet fashion)
Jeff

Last edited by Hyperveloce; 10th Aug 2009 at 18:47.
Hyperveloce is offline  
Old 10th Aug 2009, 18:11
  #4194 (permalink)  
 
Join Date: Apr 2009
Location: Petaluma
Posts: 330
Likes: 0
Received 0 Likes on 0 Posts
Proposed AD

Though the new Thales # is 'yet to be proven as robust....as Goodrich...' it stays at position 2, Goodrich on the other two stations. With a recent failure on the Thales r/r on the 320, the logic of this AD (proposed) is classic. Classically confused.
Will Fraser is offline  
Old 10th Aug 2009, 18:56
  #4195 (permalink)  
 
Join Date: Jun 2009
Location: Somewhere out there
Age: 38
Posts: 65
Likes: 0
Received 0 Likes on 0 Posts
being technical

Does this scenario of a rapid climb even require a freezing event of the Pitots ? It would be a coincidence ? (Pitot freezing event & rapid/uncontrolable climb)
Technical, accurate and objective.
augustusjeremy is offline  
Old 10th Aug 2009, 21:24
  #4196 (permalink)  
 
Join Date: Jun 2009
Location: Oxford, England
Posts: 297
Received 0 Likes on 0 Posts
Or, what happens if the protections kick out and a pilot with VERY GOOD skills
gets handed a compromised aircraft in the midst of moderate to extreme turbulence and with instrument readings which are misleading, at best? Nothing at all good I fear.
After reading this thread for several weeks, that's about the best summary i've seen to date.

I'n not qualified to comment as a commercial pilot and it's some years since I did any flying, but it amazes me that there are not more events with systems that degrade to such a condition and just at the time when the crew need them the most.

One could say not smart enough by half...
syseng68k is offline  
Old 10th Aug 2009, 22:10
  #4197 (permalink)  
 
Join Date: Jun 2009
Location: NNW of Antipodes
Age: 80
Posts: 1,330
Likes: 0
Received 0 Likes on 0 Posts
Current Data - Crash Location?

Hyperveloce;

Careful examination of the OSCAR/NOAA surface current data provides 065°T x 22.5cm/sec at 3°N 31°W over the 5 day period centered on 2 June 2009. I have therefore examined the positions in which bodies were recovered from on 6/7/8 June and constructed a likely current line based on what we know, i.e. that for the 3 days just mentioned and a calculated rate for the 5 - 6 June of 19cm/sec (9NM/day) back to the time of the accident at about 02:14:30Z on 1 June of 22.5cm/sec (10.5NM/day).

The reason for using the bodies as a check on the current is that they will have initially sunk to a point of equilibruim, and provided the depth was not too great, the water temperature would have commenced the decomposition process. Then over a period of time each of these bodies would have gained enough buoyancy to become visible on or near the surface - which explains the number of days it took to find those that they did. The point is that the bodies will have been subject to little or no leeway effects due to the surface wind. SHOM data shows that large easterly vectors on the surface become small westerly vectors the deeper you go, which helps to explain why some debris items floating with possibly little or no windage have been found to the east of the general drift line in which the bodies were found.

The reduced size graphic below shows 2 significant cumulonimbus cells, the one on the track and another left of the track shortly after passing ORARO. It seems that each of these mesoscale cells has played a part in this incident.



I surmise that for some unknown reason the WX radar has not revealed the presence of the cell the a/c penetrated at around 0209, but when everything turned pear shape at 0210 the PF made a decision to get out of the ITCZ and commenced a lefthand 180 and descent hand flying the a/c with somewhat degraded control systems provided in Alternate/Direct law. Singpilot described his experience a few posts back, but the conditions in the cockpit of AF447 were surely somewhat different.

The lefthand turn was unfortunately taking the a/c toward the Cb cell NNW of ORARO.

What happened during the SATCOM outage between 0213 and 0214 is of course speculative, but at some point in this rapid descent it can be assumed that IAS became available and an effort was made to stabilize the rate of descent. If a nose up attitude was adopted, the updraft associated with the next Cb cell may have resulted in a flameout of both engines.

Well the graphic shows the general idea, but if the current vector at 3°N 31°W was in fact 055°T x 20cm/sec, the impact point would have been about 10NM further east. This would give better GS but with a tighter turn - to be expected if the speedbrakes were deployed.

Here is the link to the full
scale graphic.

http://i846.photobucket.com/albums/a...-lkp-lge-3.jpg

mm43

Last edited by mm43; 11th Aug 2009 at 21:22. Reason: changed links
mm43 is offline  
Old 11th Aug 2009, 00:18
  #4198 (permalink)  
 
Join Date: Jan 2008
Location: Herts, UK
Posts: 748
Likes: 0
Received 0 Likes on 0 Posts
What the guy describes is basic bad stall behaviour whereby the A310 raises its nose when approaching a stall and flies itself deeper into a stall.
Somebody's going to have to be a lot more specific before this rumour is taken seriously..

Under what configuration, trim, a/s and MNo did the suggested behaviour manifest?

Then we might be able to establish how relevant it is to this thread, if it is a correct & thoroughly objective assertion.
HarryMann is offline  
Old 11th Aug 2009, 02:58
  #4199 (permalink)  
PJ2
 
Join Date: Mar 2003
Location: BC
Age: 76
Posts: 2,483
Received 0 Likes on 0 Posts
mm43;
If I may enquire, for the mathematically-challenged among us, how the radius of the turn was calculated? I would just like to understand as my very amateur reading and use of the formula from Kermode's "Mechanics of Flight" is a nominal six and three-quarter miles for a 25deg bank turn at 463kt, (532mph) groundspeed and a time of about 330 seconds for the full 180. Hand flown the turn could be made tighter of course. Speed brakes do not tighten a turn, btw, bank angle does however.

r= v^2/(g.tanθ), where;

radius is in feet, velocity is in fps, g = 32.2, Tan of 25deg = 0.4663

Found a better formula from one of our bretheren in Tech to plug into Excel:
=ROUNDUP(POWER(D2,2)/((TAN(F2/(180/PI())))*68625),3), apparently used by PAN Ops, (Link),
D2 = TAS value
F2 = Angle of bank.
Result is in NM


If the turn began at the "LKP", the yellow circle describes an 81,000ft diameter circle, or roughly 13.5 nm.

Last edited by PJ2; 11th Aug 2009 at 03:59.
PJ2 is offline  
Old 11th Aug 2009, 05:30
  #4200 (permalink)  
 
Join Date: Jun 2009
Location: NNW of Antipodes
Age: 80
Posts: 1,330
Likes: 0
Received 0 Likes on 0 Posts
PJ2;

If I may enquire, for the mathematically-challenged among us, how the radius of the turn was calculated?
Didn't even try and calculate the radius of the turn - these guys were on a roller coaster with combination of ascending / descending air and an inherited tail wind. As I said, if the speed brakes (meaning to trade speed for a higher bank angle and less Gs) had been used they would have pulled a tighter turn, but I have no idea how awkward hand flying might be in alternate/direct law - and in these conditions.

A more direct path to the calculated impact point is likely as it provides for a steeper descent and ultimately a slower IAS when they attempt to level off. Someone might like to comment on how effective the rudder would be in direct law at near stall speed pulling TOGA with only one engine. Remember, the BEA stated the vertical stabilizer damage showed the tail was rotating to port on impact and I couldn't help but wonder if only No.1 was operational.

As mentioned in the earlier post, the calculated impact point could well be 10NM further east which would fit in well with the bank angle you have indicated and what appears to have been a rapid descent. Subject to the OSCAR/NOAA surface current data being reasonably accurate, I wouldn't be too supprised to find the "Pourquoi pas?" somewhere close by.

BTW,
your circle is tangental to 0209, but moving it on 30 secs would allow the circumference to pass through the LKP at 0210. Thanks for the bank angle calcs and link.

mm43
mm43 is offline  

Thread Tools
Search this Thread

Contact Us - Archive - Advertising - Cookie Policy - Privacy Statement - Terms of Service

Copyright © 2024 MH Sub I, LLC dba Internet Brands. All rights reserved. Use of this site indicates your consent to the Terms of Use.