Go Back  PPRuNe Forums > Flight Deck Forums > Tech Log
Reload this Page >

AF 447 Search to resume

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

AF 447 Search to resume

Old 26th May 2010, 05:53
  #1181 (permalink)  
 
Join Date: Jun 2009
Location: NNW of Antipodes
Age: 81
Posts: 1,330
Received 0 Likes on 0 Posts
Originally posted by Chris Scott ...
PS: mm43, why would the aeroplane be stalled early-on, when the ASIs were UNDER-reading?
Machinbird has countered the question. In fact, I am just playing "devil's advocate" to draw out how the a/c could impact the way the BEA have reported, and within say 20NM of LKP.

In an area of 20NM radius from LKP, 40 percent of the northern semicircle has already by covered by sidescan sonar.

Regarding the ACARS messages, the timing is rounded to the nearest minute in which they were generated and is incorporated in each message prior to its identification groups. Check through them carefully and you will see that they are not necessarily in order, and those received in the first 95 seconds could have been initiated or causal to the A/P disconnect.



The above list of warnings will be displayed in HTML format if you click on the image. Bookmark the new page for easy reference - its also interactive, allowing you to highlight a line or multiple lines by clicking on them. Clicking each line again will remove the yellow highlight.

mm43

Last edited by mm43; 28th May 2010 at 02:19. Reason: added interactive feature to HTML version of graphic.
mm43 is offline  
Old 26th May 2010, 08:30
  #1182 (permalink)  
 
Join Date: Jan 2008
Location: Blighty (Nth. Downs)
Age: 77
Posts: 2,107
Received 4 Likes on 4 Posts
Quote from Machinbird:
"Chris, why do you think under-read? Is it just because all the pitot tubes are blocked?
This could be a crucial point."

Yes. Simplistically, the only way I can think that ice could cause OVER-read would be the combination of blocked static ports, but pitots NOT blocked, AND then a descent. We know they descended (though not their initial reason for doing so), but the icing combination is improbable. Does that make any sense?

mm43,

Thanks for pointing out the ACARS timing uncertainties (lack of seconds in the labels). All the events I mentioned are labelled 0210z, I think? But the order and spacing are uncertain, and I hope no-one will take my piece as purporting precision.

Chris

Last edited by Chris Scott; 26th May 2010 at 12:18. Reason: Correction of error in para 2: "climb" changed to "descent".
Chris Scott is offline  
Old 26th May 2010, 09:09
  #1183 (permalink)  
 
Join Date: Jul 2009
Location: UK
Posts: 134
Likes: 0
Received 0 Likes on 0 Posts
It has been discussed before - Pitot tubes have a drain hole, to allow ingress of liquid to drain, their calibration assumes the leak. If the drain hole blocked they will over-read, not sure what percentage. If the pitot inlet then blocked you can imagine this high reading being locked in - until the heaters melt the trapped ice in the drain and speed reading drops to zero. If the a/c flies into a fine ice cloud, all sensors could be affected in a common-mode fault. If the auto-thrust reacts to the over-read before rejecting the readings - the pilots could be handed back control with insufficient thrust to maintain current true airspeed.

The drain holes are designed to cope with ingress of low altitude rain/hail storms, and their detail design will differentiate different manufacturers models.

I assume regular maintenance procedures do non-abrasively clean these drain holes - dimensions critical, but why do problems appear to have started to occur years into operation? Of course there's also climate change...
sensor_validation is offline  
Old 26th May 2010, 10:31
  #1184 (permalink)  
 
Join Date: Jul 2009
Location: France - mostly
Age: 84
Posts: 1,682
Likes: 0
Received 0 Likes on 0 Posts
Rumours not news

Originally Posted by takata;#1011
Figaro articles about aviation are full of crap.
Rumours descending from BEA 'circles' (... the Navy experts have mistaken mediterranean tests for signals from AF447):
«Il n'y avait rien à trouver au sud, explique un membre de l'enquête.
«Il n'y a jamais eu d'AF 447 au sud, nous avons perdu une semaine, explique un proche du BEA.
Translation: we've lost a week for nothing
Counterrumours from Navy circles (not more than 24 hours):
«Ils ont passé 24 heures et pas une de plus sur notre zone, explique une source au ministère de la Défense.
BEA's Update 12/05/2010:
Exploration of the new search zone, which resulted from work undertaken by the French Navy, has continued at a speed that has been hampered by technical problems, which occurred during the dives carried out by the two Remus (Autonomous Underwater Vehicles).
The Navy again:
«Nous avons transmis une vingtaine de positions de l'Émeraude au moment où le signal des boîtes noires a été perçu», explique un proche du dossier.
Nous avons retiré certaines positions mais nous maintenons que certaines positions communiquées au BEA sont justes. Elles n'ont pas été explorées: le BEA préfère dire que l'armée s'est trompée plutôt que de dire que l'on a perçu le signal des boîtes noires mais que l'on ne saura peut-être jamais d'où il vient.»
Translation: We (Navy) gave BEA twenty locations, some of which have since been withdrawn, others remain valid but have not been explored.

Last edited by HazelNuts39; 26th May 2010 at 12:51. Reason: clarity
HazelNuts39 is offline  
Old 26th May 2010, 12:54
  #1185 (permalink)  
 
Join Date: Jun 2009
Location: Brazil
Age: 76
Posts: 35
Likes: 0
Received 0 Likes on 0 Posts
Lost: The Mystery of Flight 447

BBC 30 May

BBC - BBC Two Programmes - Lost: The Mystery of Flight 447

Lost:The Mystery of Flight 447
BBC2 Sun 30 May 2010 22:00 &
Wed 2 Jun 2010 00:20

Thanks to the person who posted this..
For those of you outside the UK , you can watch British TV with this useful little program

WebTV Unleashed: Watch Free Live Web TV Online

It does not need installing, just run it.
BUT, you must have VLC player installed and it must be in its default location
CProgram Files
It does not work here in Brazil if VLC is installed in CArquivos de Programas
so, repeat, VLC MUST be installed in CProgram Files
Gringobr is offline  
Old 26th May 2010, 13:37
  #1186 (permalink)  
 
Join Date: Jul 2009
Location: Not far from a big Lake
Age: 81
Posts: 1,454
Likes: 0
Received 0 Likes on 0 Posts
I was looking through the old AF447 thread for information and found a relevant post by Greybeard (#3399) which I will quote.
Pitot Ice

Pitot probes have a liquid drain (bleed) hole, whose air bypass has to be accounted for in the airspeed calcuation. Clog the drain, and pitot pressure rises, giving erroneous high airspeed. Indeed, per reading in this thread, some of the earlier Airbus pitot malfunctions were with poorly manufactured drain (bleed) holes.

Add more ice, and you clog the pitot head itself, and then pitot pressure will lock, or will decrease if there is any drain opening at all.

GB
Prior incidents of A330 pitot icing have resulted in overspeed warnings as well as low airspeed warnings.

But look at this post by Jeff (Hyperveloce) on the 9th of July!! #3397.
Out of the loop ?

A contribution by Pilotaydin, on the Airliners.net forum (Pitot freezing=>overestimated airspeeds=>nose up=>stall):
__________________________________________________ __________________
I would like to share a small story about something i experienced in the sim a while back, as a demo from my instructor towards the pros and cons of fbw and envelope protection...

we "flew" through an area of icing in the sim, the probe heat function and the airbus a/c itself is designed to fly through known icing, however, that doesn't mean it can withstand anything put in its path... our pitots iced over and our airspeeds started indicating 300+ at high altitude, which is bad news, because we're passing Mmo and Vmo, so the a/c as per design pitched up.... after about 20 seconds of this, as the speed wasn't decreasing, we were actually stalling and losing altitude, and the sidestick = useless, it wouldnt let any one of us pitch down, we started a large rate of altitude loss. Even if we disconnected the a/thr system and idled or added full power, the damn nose was pitched up....we went down 30,000 feet into the water outside jfk in the sim....during that descent, nothing came up on the ecam, just the warning chimes of overspeed.......we of course didnt just sit there, it was a demo we were observing he different things going on...at one point my hand did go up towards the PRIM 1 and PRIM 2 computers...i thought maybe if i let them out of the loop, we could go to altn law but i decided not to intervene to see the outcome....
knowing your systems helps, and the a/c doesnt always provide an answer to us....

Things need interpretation and over automation sometimes leaves us out of the loop...the other day over the atlantic, at 35,000 feet, we got master caution chime that said :
Start valve open and it asked us to switch off the bleed to one engine....leaving us with only one bleed left over the atlantic....are you gonna follow the ecam? or are you going to emergency cancel it?
__________________________________________________ ___________
btw, I don't get how a Pitot obstruction only (of the ram port/the drain) can lead to overestimated airspeeds (and possible overspeed alarms): if the dynamic pressure cannot be overestimated (?), then it has to be the static pressure which is underestimated (?) and this would imply that the static ports are also blocked and that altitude has been lost since their blockage ? Is that so ? Would there be other ways to overestimate airspeeds ? On the contrary, a blocked Pitot ram port is sufficient to roll back the airspeed to 70-80 kts.
Jeff
Makes the hair on the back of my neck stand up!
Machinbird
Machinbird is offline  
Old 26th May 2010, 14:30
  #1187 (permalink)  
 
Join Date: May 2010
Location: EDDS
Age: 54
Posts: 27
Likes: 0
Received 0 Likes on 0 Posts
btw, I don't get how a Pitot obstruction only (of the ram port/the drain) can lead to overestimated airspeeds (and possible overspeed alarms)
ram port + drain blocked during climb means (almost) constant ram pressure together with decreasing static pressure, leading to overreading of airspeed. Experienced it once in a Piper Seneca (FIKI approved) with malfunction of pitot heat (it worked during ground check before and after, but heating power was obviously low).
AndiKunzi is offline  
Old 26th May 2010, 14:44
  #1188 (permalink)  
 
Join Date: Mar 2002
Location: Surrey, UK
Posts: 897
Received 10 Likes on 6 Posts
If the French Navy really did confuse their own tests in the Mediterranean with the AF447 results, everyone involved would have had to overlook a difference in depth of several thousand feet, no, and of many degrees of longitude in all the locations? It's not even as if it's an obvious sort of mistake like a reciprocal or a decimal point error. Also, rather than providing "une zone qui n'etait pas la bonne", it would surely have provided invalid data.

I find this story unconvincing, and suspect the validity of Le Fig's source.
steamchicken is offline  
Old 26th May 2010, 15:59
  #1189 (permalink)  
 
Join Date: Jul 2009
Location: France - mostly
Age: 84
Posts: 1,682
Likes: 0
Received 0 Likes on 0 Posts
locating the ULB's

Just thinking aloud ...
The navy has said that it indicated twenty-odd locations where the pinger sounds might have come from, minus a few scrapped since.
The navy has also said that it identified two pingers, some distance apart.
Suppose the 20-odd locations mean different locations of the submarine, would the distance between the pingers and the difference between the distances of each pinger to the sub offer a means to locate them?

In other words, suppose a pulse from pinger A at distance sA from the sub is received at time tA, and that of pinger B at distance sB at time tB, would the difference (tA - tB) change as (sA - sB) changes for different locations of the sub?

HN39
HazelNuts39 is offline  
Old 26th May 2010, 16:40
  #1190 (permalink)  
 
Join Date: Jun 2009
Location: Chesapeake Bay
Age: 79
Posts: 57
Likes: 0
Received 0 Likes on 0 Posts
SteamChicken: I doubt we will ever know many details about the Émeraude's data and collection methods: how they code location/depth/speed of the sub and other data within or alongside an acquired signal (like SMPTE time codes to sync audio to video/film frames); when/if analog/digital conversion takes place; what and what types of filters are in place, etc. etc. Maybe somebody simply threw the wrong mag tape into his briefcase before getting on the plane, and caused some confusion. Whatever the details, it seems the Navy is rightly pissed that after a great deal of post-processing work, they proposed several search areas that were blown off by the BEA. Would be nice if folks could get on the same page and move forward. Who is leading this effort, anyway? Seems in an earlier post here it was remarked that the fellow hasn't been heard from since... In another part of the ocean, BP's CEO is visibly on the front line, despite the heat.

GB
GreatBear is offline  
Old 26th May 2010, 17:19
  #1191 (permalink)  
 
Join Date: Jun 2009
Location: in a plasma cocoon
Age: 53
Posts: 244
Likes: 0
Received 0 Likes on 0 Posts
Machinbird, I think that the past Pitot freezing events (interim report #2) tend to show that the false stall alarm rate is far higher than the false overspeed alarm rate ? beyond AndiKunzi's mechanism, maybe another one leading to (peaks of) overestimated AS was put forward some time ago (I don't remember who suggested this): when the heater eventually takes over a weakening ice crystals event, the accumulated ice in the Pitot tubes could melt then approach boiling temp creating bubbles/overpressure peaks on the analog to digital measurement device ?
HN39, if the ranges (or the different times of arrival of the pings) cannot be measured (only the directions of arrival), it relies on several triangulations from several different positions of the sub ? If a doppler (relative speed of the sub versus target) could be measured, it would provide some other angular clues.
Jeff
Hyperveloce is offline  
Old 26th May 2010, 18:40
  #1192 (permalink)  
 
Join Date: Jul 2009
Location: Not far from a big Lake
Age: 81
Posts: 1,454
Likes: 0
Received 0 Likes on 0 Posts
Jeff, I've spun the airspeed completely around climbing over weather with a frozen up pitot in a Cessna back at the dawn of time. Aircraft climb may factor in to extend the duration or degree of overspeed indication.
The relative frequency of overspeed/underspeed indications during pitot freeze up is not the issue. The issue is that it can go either way and what happens when the airspeed indications increase significantly above actual. From last years posting, it would appear to be downright ugly. I really don't think this issue has been put to bed.
Machinbird is offline  
Old 26th May 2010, 19:11
  #1193 (permalink)  
 
Join Date: Jun 2009
Location: NNW of Antipodes
Age: 81
Posts: 1,330
Received 0 Likes on 0 Posts
iabel; Thanks for your well thought-out analysis of the Emeraude's pinger detection possibilities. The locations you have indicated are ideal possibilities.

Hyperveloce; As the Navy have apparently reported that two pinger signals were detected, I wonder how they came to that conclusion. Consider that the Emeraude was moving at say 10 knots (I think its search speed is mentioned in one of the Interim Reports), then there are doppler shift events to be taken into account. These signals could possibly be indentified as separate entities if each had a detectable difference in their precise frequency and the sub's distance and speed in relation to each pinger was the same. But at this stage there is no way of knowing that, and one or both of the pinger signals detected could be multiple path. The fact that they have clearly indentified 37.5kHz pinger bursts is important, and though there is some current disagreement between the BEA and Ministry of Defence, I am sure that when Phase 4 is being drawn up, a prime area will be in the SW quadrant from West to a position WSW and between 30 to 40NM radius of the LKP.

mm43
mm43 is offline  
Old 26th May 2010, 20:28
  #1194 (permalink)  
 
Join Date: Feb 2008
Location: Europe
Posts: 80
Likes: 0
Received 0 Likes on 0 Posts
The BBC blurb about the TV show says:

One year on, a full explanation of what might have happened has emerged
which sounds dubious, especially as the producer was ringing round the houses looking for a sim. Hope it's not going to be long on drama, short on sense.
GobonaStick is offline  
Old 26th May 2010, 21:07
  #1195 (permalink)  
 
Join Date: Nov 2001
Location: UK
Age: 70
Posts: 76
Likes: 0
Received 0 Likes on 0 Posts
I’m curious about this BEA position of not searching to the south.
Is there something in the ACAR’s transmissions that the BEA have not fully explained.
Now I have no expertise in this area, but is it possible that the transmit time and receive time get shorter (by milliseconds) assuming the satellite is to the relative north of transmission, so if the aircraft turned south the transmit/receive time would increase. Someone here must know the answer to this, if not shoot me down in flames for being naive.
Backoffice is offline  
Old 26th May 2010, 21:08
  #1196 (permalink)  
 
Join Date: Nov 2009
Location: Near Puget Sound
Age: 86
Posts: 88
Likes: 0
Received 0 Likes on 0 Posts
Airspeed reading 300+ -- shades of Northwest B727 12/1/74

BTW, this accident to a feloow NW pilot got me started in the whole area of aviation human factors and accident investigation.


Goldfish
goldfish85 is offline  
Old 26th May 2010, 21:29
  #1197 (permalink)  
 
Join Date: Dec 2005
Location: No. Cal, USA
Age: 72
Posts: 112
Likes: 0
Received 0 Likes on 0 Posts
Forgive the slightly off-topic inquiry. Has there been any statement from Airbus or BAE regarding similar sequences of ACARS messages from other planes that might have suffered similar upsets?
grumpyoldgeek is offline  
Old 26th May 2010, 22:08
  #1198 (permalink)  
 
Join Date: Jun 2009
Location: NNW of Antipodes
Age: 81
Posts: 1,330
Received 0 Likes on 0 Posts
originally posted by Backoffice ...
... is it possible that the transmit time and receive time get shorter (by milliseconds) assuming the satellite is to the relative north of transmission, so if the aircraft turned south the transmit/receive time would increase
The satellite is geostationary over the equator and in theory the the path length will get shorter when proceeding south toward the equator. However, without highly accurate timing markers in a transmitted stream of data, there is no way of measuring the path length. Even if the satellite had initiated timing, a return signal to it would have passed through indeterminable delays in the aircraft receive/transmit path.

Many moons ago, I did check which satellite was being used, but can't remember whether it was east or west (possibly a few degrees east) of the LKP.

A good thought, but not practical in this case.

mm43
mm43 is offline  
Old 26th May 2010, 22:13
  #1199 (permalink)  
 
Join Date: Jun 2009
Location: NNW of Antipodes
Age: 81
Posts: 1,330
Received 0 Likes on 0 Posts
originally posted by grumpyoldgeek ...
Has there been any statement from Airbus or BAE regarding similar sequences of ACARS messages from other planes that might have suffered similar upsets?
Yes, and in recent times there was a QF72 upset that was determined not to be relevant, and others may have been similar, but deemed not to be the same.

mm43
mm43 is offline  
Old 27th May 2010, 02:02
  #1200 (permalink)  
 
Join Date: May 2010
Location: MA, USA
Posts: 126
Likes: 0
Received 0 Likes on 0 Posts
Originally Posted by mm43
iabel; Thanks for your well thought-out analysis of the Emeraude's pinger detection possibilities. The locations you have indicated are ideal possibilities.
Thanks for the reminder that I had not studied the post from iabel. It is an interesting analysis, pulling together aspects that we have discussed separately. I would caution that the distance from the center of the Navy search box to the mountain range is 12km; add to this the slant distance from the pinger to the mountains. 12-18km is a long way for a weak 37kHz signal, but perhaps not impossible with the "right" focusing and ducting.

Originally Posted by mm43
Hyperveloce; As the Navy have apparently reported that two pinger signals were detected, I wonder how they came to that conclusion. Consider that the Emeraude was moving at say 10 knots (I think its search speed is mentioned in one of the Interim Reports), then there are doppler shift events to be taken into account. These signals could possibly be indentified as separate entities if each had a detectable difference in their precise frequency and the sub's distance and speed in relation to each pinger was the same. But at this stage there is no way of knowing that, and one or both of the pinger signals detected could be multiple path. The fact that they have clearly indentified 37.5kHz pinger bursts is important, and though there is some current disagreement between the BEA and Ministry of Defence, I am sure that when Phase 4 is being drawn up, a prime area will be in the SW quadrant from West to a position WSW and between 30 to 40NM radius of the LKP.
I was thinking that reception of two pings per second would be sufficient evidence of two pingers, but I agree that the effect could be mimicked by multipath. I expect that frequency is the key.

The Doppler shift (if I have the math right), based on 10kt=5m/s, is 5(m/s) / 1500(m/s) * 37.5(kHz) = 125Hz. That is the maximum possible (not counting for currents) and the actual is likely much less if the sub only hears the pinger when the pinger is below or to the side of the sub (range-rate = speed * sin(angle from vertical) * cos(angle from bow)). However, if the sound was ducted or refracted to near horizontal, then the full Doppler could be detected if the target was ahead or astern.

Note, however, that the frequency of the pingers is badly controlled, likely by an RC timer. The spec says the frequency is 37.5 +/-1kHz. That leaves plenty of room to separate one pinger from another, even with Doppler obscuring the result, unless the two pingers happen to be close in frequency.

As for exploiting the Doppler shift for localization, that is certainly possible in principal, but may be difficult with the amount of data they have. The pulse length is only 10ms, which translates to 375 cycles. The maximum possible Doppler shift is 5/1500 = 1:300. Thus the Doppler has to be detected from about 1 cycle or less change in length of a signal buried in noise. If the aspect isn't favorable, that may be hard to do. Someone who has experience with Doppler measurements will know better. Doppler is routinely used to measure water current from the weak echos reflected from particles and turbulence in the water, but that is done at 10-50 times higher frequency, receiving with the same clock that was used to transmit, and averaging over multiple pings.
auv-ee 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.