Air France jet missing over the Atlantic

[Phil Bunch]

Hello,

I have been following this story with interest and now the sad resignation that we must accept the worse case.

I do have some questions that have come up:-

In terms of the black box, what is the media inside that is used to record the data?
Assuming that this aircraft was lost where the few parts have to date been found, what is the surviability of the black box? I mean we are in area where the sea is potentially 7000m deep from what I have read? So does the black box stay afloat? Does it have a transmitter that can assist in locating it? If so what is the range of the transmitter?

As well as sensors from the aircrafts systems, does the black box have sensors to give independant readings e.g., G forces, altitude etc.?

A further thought - Is this the farthest away from land that an aircraft has been lost?

Regards

Carl

==============

Carl,

Below is an excerpt from a recent online news article, and it contains some information regarding the detection of the black box recording devices underwater, range of detection, etc.

http://online.wsj.com/article/SB124394381881876429.html

By my own guesses and informal research, the depth of the ocean in the region of the debris field is so deep (over 10,000 feet) that recovery of the black box will be very hard.  I am guessing that much will depend on how accessible the recording boxes are and how fragmented the aircraft's pieces are.  Previous mid-air explosions over the ocean have yielded some fairly large segments with a huge number of small pieces.  

The data and cockpit voice recorders could even be buried in a complex pile of tangled wreckage.  Based on watching a few TV shows, undersea robots can do a limited amount of underwater salvage and recovery at these depths IF they can find the black boxes before their underwater beacons run out of batteries in about 30 days.

News article excerpt:
---------

The fastest path may be focusing on the recorders' emergency signal, said Barry Cairns, a sales manager at Sonardyne International Ltd., a company outside London that specializes in undersea navigation and positioning, largely for the energy industry. Sonardyne in 2006 helped retrieve the black boxes from an Armenian Airbus jetliner that crashed in the Black Sea.

Typically, searchers drag a sensor from a ship that traces a grid on the surface. The smaller the search area, the faster the sound is likely to be detected. Once the signal is heard, computers can pinpoint its source quickly, Mr. Cairns said.

Undersea, the signal can carry for roughly 3,300 feet, so "it all depends on how deep the black boxes have gone," he said.

One factor that makes this accident so unusual is the remoteness of the apparent crash site and the depth of the water there, which could exceed 13,000 feet, roughly the depth of the Titanic.

-----------------------

[Phil Bunch]

An article on the Air France crash in a UK newspaper this time.  

They attempt to make connections between a recent Australian incident in which the navigational electronics became very erratic, causing the airliner to fly very erratically.

http://www.guardian.co.uk/world/2009/jun/02/france-plane-crash-airbus-system/print

I don't get why there is such a tendency to blame modern fly-by-wire/computerized airliners - the accident rate is at an incredibly low rate now due to the superiority of these software-based systems.  While no such system will ever be perfect and nothing this complex can possibly be perfect under all circumstances, these are surely among the most reliable systems mankind has ever created.  The accident and incident rate is so low it is statistically all but impossible to improve them, IMO.  I don't know of any statistics supporting that Airbus is less reliable than other vendors, and the number of accidents is so low that any numerical differences in accident rates would probably be statistically insignificant.  The number of events is just too low to use statistics, and without statistics it is very hard to say anything meaningful about such things.

Below are some excerpts from this article.  It's apparent that the lawyers have as usual become involved, hoping to obtain their usual large sums of money from the airlines and the aircraft companies.  I assume the suit will be files in the USA with our unfortunate history of huge jury awards for damages, etc.  Unfortunately, in the USA truth is not too much of an issue - only the ability of the lawyers to convince a jury of mostly poorly educated ordinary citizens.  Typically, the companies being sued just settle out of court to avoid the worst case scenarios of a jury trial.

-----------------

A law firm representing victims of a previous serious incident involving an Airbus A330 warned that the aircraft could be vulnerable to electrical interference.

London-based Stewarts Law is preparing to sue Airbus, the Australian airline Qantas and a computer manufacturer, on behalf of 30 people injured when an A330 suddenly plunged 650 feet in flight last October. Air accident investigators blamed the rapid descent at 37,000ft over Western Australia on a malfunctioning computer system. The A330 is a "fly-by-wire" aeroplane that uses an air data computer to guide the aircraft and, if necessary, amend pilot errors.

The A330 has flown 707 million passengers since its launch 15 years ago and this week's crash was its first fatal incident in active service. But in the Qantas case, the aircraft's main navigational tool, the air data inertial reference unit, started emitting "electronic spikes" and plunged the A330 into a dive after mistakenly calculating that the aircraft was pitching its nose into the air. In fact, the aircraft was flying level.

According to one theory being examined by lawyers, the Airbus computer systems could have been affected by inadvertent interference from a nearby naval communications station. Jim Morris, a senior associate at Stewarts Law, said the Air France crash had again raised fears about the aircraft's susceptibility to electromagnetic interference. "It could be that if the A330 is more vulnerable to electromagnetic interference, it could have caused the pilots to lose control of the aircraft during severe turbulence," he said. He added: "The indications from the aircraft's datalink system are that there was a loss of electrical systems in a very short space of time, which indicates that there was a catastrophic failure."

--------------------------

[Peter Sagar]

A very good mate of mine is F/O on the A330 here in Australia. With respect to the ongoing problems with AIRDU units installed on the A330, apparenlty the Airbus guys are completely unsure as to what is causing the problems with uncommanded auotpilot inputs. They know what is happening, just can't pin down a specific cause. Airbus technicians at a recent briefing essentially blamed H******ll as the unit is supplied by them.

My mate's simplistic explanation of the new unofficial procedure when the A330 suffers an "upset" is to switch OFF the "flight computers" and very briefly revert to the redundancy system. Then after a short time, re-start the computers. BUT, don't quote me on that procedure, as there was a large quantity of amber fluid consumed prior to and during that conversation.  :twisted:  :twisted:

Seems to me that the B744 systems offer an older but far more stable and confident option.

Peter.

[Phil Bunch]

I read an airline magazine real life story about a US Air pilot whose trans-Atlantic airliner's navigation computers lost their way for no apparent reason during flight, part-way across the pond.  His First Officer had to manually track their position every 5 minutes all the way to Heathrow - they were advised by maintenance not to reboot the computers since it might make things worse.  After the plane landed, the maintenance staff saw that the system was not working properly but after it was rebooted they could not reproduce the error condition.  I'm not sure I would have wanted to be the next crew heading back across the Atlantic with that plane!

Intermittent errors and problems with any system can be terribly hard to diagnose and repair.  We've all had that experience with automobiles, even back in the day when they were mechanical and not rolling computer networks.  I'm not sure ANYTHING other than the steering and parts of the braking system and the wheel bearings and tires on my Volvo AWD S60 is directly mechanical.  Even the door locks are merely signals sent to a networked computer which then controls a solenoid.

As the old IBM mainframes used to say about almost every error condition:  "IHC 240I - Probable user error."  I came to really hate that message.

[Qavion]

In terms of the black box, what is the media inside that is used to record the data?

Well, an aircraft of this generation should have a Cockpit Voice Recorder(CVR) and a Dignital Flight Data Recorder(DFDR) with a solid state memory (i.e. completely shockproof). There will still be aircraft around, however, with magnetic tape recorders, but these are designed to operate under high g forces and will survive catastrophic g forces. These two types are generally interchangeable, however, so I couldn't be sure what type was fitted.

A typical black box memory will be encased in several layers of fairly thick metal (these boxes are quite heavy and definitely do not float) and may have a water jacket around one layer. In intense heat, the water turns to steam, and steam generally is limited to a specific temperature range well below the intense heat of an aircraft fire.

Assuming that this aircraft was lost where the few parts have to date been found, what is the survivability of the black box?

The two black boxes generally survive most accidents. This is partly because they are located in the tail area, which seems to remain fairly intact even in bad accidents. The boxes can be found on electronic racks with, I recall, spring and rubber, shockmounts.

There is an ultrasonic beacon/transmitter mounted on the front of most black boxes which activates on contact with water (the water provides a conductive medium for the circuits to start working) By the way, most of these boxes are painted in a fluouro orange colour with reflective stripes stuck on the side, so this does help searchers, in many cases, to visually locate and identify the boxes.
Ultrasonic transmitters are required as radio waves don't go very far underwater. The figures given for transmitter battery life are probably worst case scenario. The batteries have a "shelf life" of many years and are renewed periodically. The battery could be new.

As well as sensors from the aircrafts systems, does the black box have sensors to give independant readings e.g., G forces, altitude etc.?

Not sure about the A330, but the 747-400 has a dedicated accelerometer mounted in one of the main gear wheel wells. Considering the wreckage was reportedly spread over a very wide area in this particular incident, I'm guessing the aircraft was breaking up before it hit the water, so there is no guarantee that the accelerometer would still be working (or, indeed, the black boxes themselves).

Many aircraft also have another emergency transmitter mounted on top of the fuselage, near the tail which automatically transmits UHF and VHF signals when high g forces are experienced. However, if this particular part of the fuselage is underwater, then it will be basically useless.

Hope this helps.

Cheers.
Q>

[John Golin]

The ISIS referred to in the alleged ACARS is the standby instrumentation (and associated sensors?)... the fact it was able to send ACARS messages advising of failures seems strangely at odds with the extent of failures - to lose so much and still have ACARS sending info... very strange...

sidetrack: Qavion, does the 777 have the same Honeywell ADIRU as the 'bus?

[Jeroen Hoppenbrouwers]

... they were advised by maintenance not to reboot the computers since it might make things worse.  After the plane landed, the maintenance staff saw that the system was not working properly but after it was rebooted they could not reproduce the error condition.

Ah, normal fault-finding procedure in commercial computer operations   :twisted:

It is terribly hard to diagnose a malfunctioning machine, and there may be pressure from users and management to "just fix it" which means a reboot and, usually, throwing away all evidence of what went wrong. I've seen many cases of this, and the more important the system, the less time the technicians get to play with the malfunctioning state. Only a fully redundant system that can be unplugged without consequences can be properly diagnosed, usually.

The same situation is common when systems get compromised due to a security problem (breakin, contamination, etc.). A common reflex is to wipe the disks and re-install the backup. That destroys all evidence of what caused the compromise in the first place. But, since it is a machine that has been carefully broken in to, leaving it running also has the definite possibility of destroying evidence, as it may have been booby trapped. Unplugging the network connector has been known to trigger a self destruction software device. You'd like to "freeze" the thing instantly, which is easier said than done.

I do agree that in case of a still flying airplane, an inflight reboot should be considered a measure of last resort, because you never know that it will clear the problems.


Jeroen

[Qavion]

sidetrack: Qavion, does the 777 have the same Honeywell ADIRU as the 'bus?

Sorry, John. I currently don't have access to this information. Our own airline doesn't fly 777's and has very limited exposure to foreign and other operators these days. The 777 maintenance manual I have access to doesn't mention manufacturers or p/n's.

Rgds.
Q>

[CarlBB]

Thanks for the replies above; very interesting reading.
Rgds
Carl

[Shiv Mathur]

A typical black box memory will be encased in several layers of fairly thick metal (these boxes are quite heavy and definitely do not float) and may have a water jacket around one layer.

Okay, here's my idea. It surely can't be impossible to have another layer filled with air (or even Helium) .. the idea being that it would float to the surface of the water.

It's jammed up in other parts that won't let it rise?
Okay, since it's in the tail, it must be easy to place it in a section that breaks off on heavy to moderate impact, as surely any collision with the water would be.

If this is an old idea, sorry.
If it's a brilliant idea, make sure I get royalties, gentlemen

[Phil Bunch]

Honeywell is quoted as saying that the "black box" recorders have underwater beacons that can be tracked underwater at a depth of 20,000 feet, and even from the surface.  This is much more than the previous 3000 feet others had claimed.

Below is the URL for the updated article:

http://online.wsj.com/article/SB124394381881876429.html

an excerpt follows:
=====================
The smaller the search area, the faster the sound is likely to be detected. Once the signal is heard, computers can pinpoint its source quickly, so "it all depends on how deep the black boxes have gone," Mr. Cairns said.

Honeywell International Inc., the recorders' producer, said that barring unusual currents and temperatures, the signal should carry for at least 20,000 feet and may be strong enough to be picked up on the surface.
=====================

[Garry Richards]

In intense heat, the water turns to steam, and steam generally is limited to a specific temperature range well below the intense heat of an aircraft fire.

The purpose of the water jacket is probably to limit the rate of temperature rise in the innermost compartments as it requires much more energy to convert boiling water to steam than it does to heat it from cold to boiling. The hope would be that the intense fire has died down by then. If the fire is still bruning then once the water has turned completely to steam the temperature on the inside will resume rising until it reaches the temperature on the outside.

Okay, here's my idea. It surely can't be impossible to have another layer filled with air (or even Helium) .. the idea being that it would float to the surface of the water.

For an object to float, its density has to be less than that of water. As steel is much more dense than water it would have to have a large volume of less dense material inside it, like a submarine. The 'black box' would have to be very large, with lots of empty space within for it to be able to float. Squeezing a gas into a small cavity would not work.

[Phil Bunch]

A typical black box memory will be encased in several layers of fairly thick metal (these boxes are quite heavy and definitely do not float) and may have a water jacket around one layer.

Okay, here's my idea. It surely can't be impossible to have another layer filled with air (or even Helium) .. the idea being that it would float to the surface of the water.

It's jammed up in other parts that won't let it rise?
Okay, since it's in the tail, it must be easy to place it in a section that breaks off on heavy to moderate impact, as surely any collision with the water would be.

If this is an old idea, sorry.
If it's a brilliant idea, make sure I get royalties, gentlemen

======================
Shiv,

I have an idea,, too!   Put the recorders into rocket-assisted probes that will detach from a crashing airliner.  Shoot them up a few miles and then launch the parachutes and start the flashing strobe lights, radio beacons, GPS trackers, etc.  (insert many grins here)  

Along somewhat the same line of thought - there are rumors that Air Force One has an escape pod for the President.  No confirmation on this but it was featured in an action movie a while back.  

On a more realistic note, there has been talk of having video recorders in the cockpit, not just voice recorders.  That way I guess one could see what the heck is going on and/or what mistakes a pilot makes during crises.  I fear the recent Buffalo icing accident/stall would not have been a pretty scene on video as he fought the stick pusher...

[Qavion]

Honeywell is quoted as saying that the "black box" recorders have underwater beacons that can be tracked underwater at a depth of 20,000 feet,

My notes have a different interpretation. The operating depth of the beacon is quoted as being 20,000'. i.e. The locator beacon on the front of the DFDR/CVR is in a metal tube (about 1/8" thickness, I'd estimate). It will probably collapse at depths greater than 20,000'.
My notes also say the transmitting range varies depending on the sea water (1800~3000metres)

The old FDR weighs 30lbs, the new one 18lbs. The old CVR was 24lbs. How much helium would you need to lift this? : )

Rgds.
Q>

[Dominic Manzer]

Properties of the Flight Recorders:

I was a designer, with NASA, of flight electronics and the back boxes that enclose them for many years. An electronics box designed to withstand the typical G load and vibration environment of aircraft flight or spacecraft launch ends up being close to the density of water (Specific Gravity = 1). Some a little more and some a little less. Constructing an inclosure that can survive a 1000 G plus impact adds so much metal that the boxes' SG will increase to the 3 - 5 range.

One of the FR's from the Air Florida crash was found when a searcher realized that he was only knee deep in water when elsewhere the water was wast deep. He was standing on it. This FR was completely on it's own, not entangled with other structures, and obviously not floating.

I don't know how much effort is put into sealing flight recorders (FR) as air spaces are counter productive. Unless one goes to extreme measures when an air filled container sinks in a deep ocean it will eventually implode, pulverizing what ever is inside.

When a FR is recovered from water, the NTSB assumes that it is full of water. If at all possible, the wet FR is placed in a cooler, filled with the water the FR was found in without exposing the FR to air.

The underwater beacons produce "chirps" that are clearly audible to the human ear. People who were present when a FR was recovered from deep water often commented that they could hear the beacon. The beacons in both of Air Florida's FR's failed. I never herd why they failed, presumably due to an unusually hard impact. If a lack of maintenance or some other apparently easily correctable problem caused the beacon failure, it would have been all over the news.

Sense the "official" French deep ocean search team will not arrive on sight for several days, other "national assets" (subs) of French, US, British or evan Russian origins might be brought to bare. A sub's passive sonar might be able to hear the beacons from considerable distances perhaps a 100 miles or more. If you doubt this,  experimental transmissions have been made from the indian ocean, near Malaysia to Woods Hole, Mass and back using the "deep ocean acoustic channel." If a  "national asset" is close enough, the official search team will get a message, look hear, or don't look over there with little other explanation. This may be the real reason the US NTSB as gotten involved when there are few signs that the US manufactured engines or avionics were involved.

The signal saying loss of cabin pressure and electrical failure is quite ominous, indicating a midair break-up of the aircraft. Some people have questioned how anything is transmitted with a power failure. Transmitting this much information could be so fast that the equipments power supplies had not discharged yet. The aircrafts location dictates that the transmission had to be over the aircrafts primary communications systems, either long distance radio or satellite. If the message was limited to just those two failures, I'd suspect that the transmission was radio, with it's limited bandwidth. Normally, when a system transmits this type of cry for help, large amounts of supporting data are sent. For aircraft in flight, what's working is often more important than what is not. The reports I've herd said the transmission stopped abruptly, suggesting that normally more would have been sent.

[Qavion]

"The underwater beacons produce "chirps" that are clearly audible to the human ear. "

Interesting. I've tested these devices many times and on all occasions have had to use an electronic device to listen to the chirps. The very name of this device suggests it is beyond the human hearing range. If the Air Florida aircraft crashed in Florida(?), I'd be more inclined to think that the chirps were insects or frogs

Having said that, the beacon has not been submerged in water when I've tested it. Perhaps the water modifies the sound? Or perhaps you need to stick your head underwater to hear it?

Rgds.
Q>

[Phil Bunch]

Honeywell is quoted as saying that the "black box" recorders have underwater beacons that can be tracked underwater at a depth of 20,000 feet,

My notes have a different interpretation. The operating depth of the beacon is quoted as being 20,000'. i.e. The locator beacon on the front of the DFDR/CVR is in a metal tube (about 1/8" thickness, I'd estimate). It will probably collapse at depths greater than 20,000'.
My notes also say the transmitting range varies depending on the sea water (1800~3000metres)

The old FDR weighs 30lbs, the new one 18lbs. The old CVR was 24lbs. How much helium would you need to lift this? : )

Rgds.
Q>

====================================

Below is another article, this time with a little more technical info but also lots of speculation by various people.  For technically complex stories like this, this newspaper isn't as bad as some, but they were recently sold to a big media conglomerate and have become more sensationalistic and politicized.  The USA is losing most of its newspapers to the internet very rapidly.

http://online.wsj.com/article/SB124402430768380893.html#printMode

I thought the discussion about converting airliners to a higher-bandwidth, higher information content cockpit and other data transmission system sounded promising, but since I don't know anything about these issues it's easy to convince me!

Take  a look at this excerpt from the article below - I didn't know it was possible to send text messages through the maintenance-data transmission system(s).  Or is this capability just speculation, too?  It doesn't make sense to me that they would try to signal for help or otherwise communicate via this semi-offline maintenance data collection system in the middle of a major in-flight crisis.  I can just see someone typing a text message letter by letter instead of trying to help recover the plane.  The message might go something like this (satire of news article intended):

"Is anybody reading this message?  Radio doesn't work.  Plane is falling apart over Atlantic!  We just lost cabin pressure too!  We're going down!  Is anybody there?  HELP!!!"

What can they do, key in short text messages through their normal navigation keyboards?  Sounds worse than trying to use a Blackberry or a smartphone.  Surely they would instead use a radio if they had time to send text messages?  

Is any type of worldwide satellite phone normally available to transoceanic crews?

====================
"Investigators also will scour records to determine if the pilots attempted to send any text messages using the maintenance-data transmission system."
====================

[Pesach]

I know I appear and disappear on the aerowinx forum on occasion, and I apologize for that... :oops: i try to follow the forum whenever I get the chance to.

Anyhow, I just wanted to share the following URL that I stumbled upon an hour ago. It contains an outstanding (in my opinion) meteorological analysis of AF447's flight path:

http://www.weathergraphics.com/tim/af447/

Haven't had the chance to read through it completely yet, but it appears to be very comprehensive!!

-Pesach (KJFK)

[Dominic Manzer]

I miss-spoke. Air Florida was not the accident I was referring too, the correct name escapes me, They did crash in the Florida swamps, after expired oxygen generators started a cargo hold fire. The searchers were latterly up to their person holes in snakes and alligators. The assets of the airline were sold and became what I think is now AirTran. (I'm terrible with names.)

Low frequency sounds propagate farther in water than higher ones. Why would anyone make the sound harder to detect and fade away quicker by making it to high too hear?

Depending on the design of the transducer the presence of water may greatly effect the operating frequency. Water is 4? orders of magnitude denser that air and the acoustic loading of the transducer will be that much different, potentially shifting the resonance frequency from 100,000 Hz to 10 Hz, from well above human hearing to below. While the actual frequency shift is probably much less, a shift from ultrasonic (above 20k) to the hearing midrange, 2 to 3k in water is quite likely.

If the transducer is capable of wide band operation, testing it at high frequencies may be simply to avoid annoying people not involved with the test. In your environment, ware testing is only occasional, the sound would be of little concern, but in the factory that makes them, the sound could be a real problem. Often details of test procedures used in production persist well after they have any meaning out side of the plant.

Also, I believe the batteries for the pinger are water activated. Water activation removes nearly any possibility of the battery discharging before it is needed. The only limitation on the shelf/service life is moisture absorbed or condensed from the air. Of course condensation can be a serious problem for aircraft. Having the battery water activated simplifies and improves the reliability of the system by removing the need for some kind of switch to turn the thing on.

[John Golin]

sidetrack: Qavion, does the 777 have the same Honeywell ADIRU as the 'bus?

Sorry, John. I currently don't have access to this information. Our own airline doesn't fly 777's and has very limited exposure to foreign and other operators these days. The 777 maintenance manual I have access to doesn't mention manufacturers or p/n's.

Rgds.
Q>

I should have realised your exposure would be limited given the last few years of 'management'  Thanks anyway.