Mixed Signals
Any sufficiently advanced information is indistinguishable from noise
Monday 21 January 2008, 1:14 PM
BA038 crash at Heathrow
As air accidents go, this one is nearly perfect. A Boeing 777 falls out of the air, not in itself a good thing, but chooses to do so on just about the only patch of soft ground in London surrounded by aviation-trained fire crews, accident investigators and nice people in uniform proffering hot tea. Everyone on board escapes serious hurt and walks – or at least hobbles – to safety, BA get to parade a Hollywood-perfect crew in front of the world's media, and a few days later the wreckage is quietly tidied away.
There are just a couple of points outstanding, one human, one technical. The human side is easy: by all accounts, the crew really do deserve those medals the size of frying pans. There's an old RAF saying – if it looks like you're due a prang, select the softest and cheapest object in the landscape and endeavour to strike it as gently and slowly as possible. The lawn in front of Runway 27L qualifies perfectly, and the pilot settled the aircraft down from a few feet up by stalling it. He did this after making all the right decisions and exhibiting exemplary flying skills in a situation thought impossible and thus never trained for – with no time for any sort of planning nor margin for error.
The technical side is a bit more pressing. By now, the Air Accidents Inquiry Board (AAIB) has determined that both engines went on strike, sulking at flight idle when asked for more power, Scotty. They're not supposed to do that, you know. Losing one is unfortunate, two begins to look like carelessness, but there's so much duplication in a modern two-engined passenger jet that a common cause is very hard to find. Fuel contamination, icing, pilot error, downdrafts and bird strikes have all been effectively ruled out – at least by the best informed guesswork; the AAIB may know more but it's not saying yet.
And as Sherlock Holmes said, with bad logic but good rhetoric, once you've ruled out the probable whatever's left, however improbable, is the cause. That means that the armchair engineers, conspiracy theorists and axe-grinders are coming out to play – none more so than with the hertzian monsters who live in the Land of Electronic Smog.
An early example of this – expect many, many more – comes from one Nina Anderson, author of the spine-chilling tome 'Worse Than Global Warming – Wave Technology': "According to Anderson, wireless technology is rampant and the unseen hazards from this technology could influence our health, behaviour and maybe aircraft controls."
Yeah, and gremlins on the wing drove the Captain mad.
Which is not to say that radio frequency interference might not be a factor. I think it's very unlikely, but there is anecdotal evidence and some test results that suggest under exceptional conditions that prove difficult or impossible to replicate, avionics can be affected by personal mobile equipment.
Given the phase of the flight when the accident happened, the primary suspect for the source of any such interference has to be mobile phone handsets, illicitly turned on by business travellers desperate to make contact after a long flight and connecting with base stations as they finally come within range.
And if this is the case -- again, it's a huge if -- there'll be plenty of evidence in the logs of the base stations themselves. The mobile phone operators will have access to the details of which handsets made contact with what base stations when, and the signature of someone connecting out of the sky in the approach to Heathrow will be distinctive enough that pattern-matching searches should produce a list of suspects with little effort. It may be worth looking at this sort of information in general, to work out how often it happens. Then, by matching the handset owner records with the flight manifest, the accident investigators can get the actual equipment used and do further testing.
That's what solves cases like this; good old fashioned sleuthing rather than calling up invisible mysteries and tying them to weird health scares. And in the AAIB, we have some very good old fashioned sleuths dealing with one of the best-instrumented airframes in the world, with all the evidence, physical, electronic and human, intact and available. You couldn't hope for a better accident.
Worse than global warming. Sheesh.
PreviousComments on this post
There was a long period of uncertainty and controversy when the possibility of using twin-engine aircraft for trans-Atlantic commercial flights was first brought under consideration. As you say, putting the plane down gently on the nearest, softest piece of ground, is a good solution and may involve heroic actions. Having the same thing happen when there is no soft ground within range -or any ground at all, for that matter - can end tragically, despite any number of heroics.
If the cause does turn out to be loss of engine power, lack of response to throttle commands, it will be interesting to see what the ramifications are.
What about the cell phone jamming equipment that would have been accompanying Prime Minister Browns convoy at the airport. Not likely a single cell phone caused this, but perhaps jamming equipment that covers all the spectrum. The computer didn't deliver the messages to the engines in my opinion because the signal was stopped by the jammer.
ian
That's highly unlikely, Risky. The Prime Minister's vehicle was a long way away from the incoming BA flight - the newspaper reports were incorrect - and the ECM is designed to be as undisruptive as possible.
In any case, broad spectrum jamming wouldn't be the problem (it might affect navigation equipment, but that's not at question here). It would have to be high levels of RF.
R
I find it hard to believe the engines ran out of fuel, the BA would not have been so quick to label the pilots heros. The engines failed to spool up ( all of us are familiar with that sound ) as we approach landing. The pilot pushing on the throttle is no different than the computer pushing on the throttle, it is the same interface that connects to both engine controls . The fact that they didn't respond, tells me that neither engine computer was getting the signal from neither of the interfaces, practically unheard of..well...unheard of.
I also find it hard to believe a simple cell phone on a plane would do it, but that will be tracable anyway, so they may yet find out what happened.
I understand your commet re broad spectrum jamming, but there is also such a thing as upper and lower sidebands in any set of telecommunications frequencies, so any combination of any other frequencies in the vicinity at the time could have become low level RF and interfered with the computers. The ECM may be designed to be as undisruptive as possible, so if that was the case, why ban cell phone operation? They have about .6 watt output of RF. Also the fact that no other 777 flights were grounded tells me that they have a good idea of what happened already . Hoepfully it is something simple.
Thanks for the opportunity to add my 2 cents.
ian
Whatever the relative power levels of the PM's ECM and mobiles on board, both would have been obliterated by the landing radar.
Which prompts the question, why was the pilot landing manually? I was under the impression that all flights into Heathrow were brought down by computer.
Chris, he only switched to manual when the engines failed to spool up , and he quickly grabbed the controls to dip the nose and increase speed to avoid a stall, then raised the nose and belly landed perfect when he was in the landing area....just a few feet short, but a soft landing as opposed to a concrete runway.
I believe it's common to manually land aircraft. As a pilot, you have the discretion to leave as much or as little of the flying to the computer as you like, and on a 13 hour flight from China I'd be surprised if much of the cruise phase was hand-flown! Landing is another matter, and from what I understand auto-land is kept as a last resort for airfields which are properly equipped and in weather (like heavy fog) when it's the safest option. Normally, the landing is by hand, because it's a situation where things that go wrong need immediate action. The pilot flying takes over from the computer at various points depending on weather and other factors.
In this case, I don't know whether the pilots were steering the thing during finals - but it was on autothrottle, with the computer adjusting thrust to keep it on the glidepath. When that failed, the pilots became aware that the autothrottle wasn't working and tried to set the thrust by hand - and when that failed, they realised that things were going badly awry and there was no question of automation thereafter.
As for landing radar - not quite sure what you mean by that. The only radar civilian aircraft carry is for weather and that'll be turned off at this point; there'll be radar at LHR, but nothing devoted to landing alone. In any case, the avionics will be specifically designed to cope with radar emissions (indeed, some are designed to receive and react to them), because the designers will know about those, their power and frequencies.
Passengers' mobile devices are far more of an unknown, and while each of them may be under a watt there's a cumulative effect - and they're inside the Faraday cage of the aircraft, while external sources are shielded. It's very hard to come up with an approval regime that can cope with so many variables, so the safe choice is to put a blanket ban on transmitters.
As for sidebands - they're a part of all radiocommunications, but they only happen when you mix two frequencies. That requires quite a specific set of circumstances, which normally only occur within a radio receiver (which can be expected to have quite a lot of filtering to get rid of the unwanted mixing products). Two radio carriers coinciding in free space - or when impinging on metal - won't create sidebands.
This is real interesting, and although it refers to GE engines, the results could be the same for the RR engines :
here is the link, but scroll down to the bottom AD.
http://www.iasa-intl.com/folders/belfast/BA777_Unthrustworthy.htm
---------------------------------------
The Original AD for General Electric's 777 Engines
Granted the following AD applied to GE90's but does appear to confirm that there are paths where FADEC signals can be corrupted.
SUPPLEMENTARY INFORMATION: The Federal Aviation Administration (FAA) has received seven reports of loss of thrust control (LOTC) on General Electric Company (GE) Model GE90 turbofan engines installed on Boeing 777 series aircraft. Five LOTC events occurred in-flight and two occurred on the ground. The five in-flight LOTC events were temporary in that the engine recovered and continued to operate normally for the remainder of the flight.
Investigation
The investigation revealed that water can accumulate in the Ps3 and P3B pressure sensing system, which can freeze in the full authority digital engine control (FADEC) sensing ports or pressure line. Frozen water can result in a restriction or a blocked signal to the FADEC. This blocked signal can cause a corruption of the FADEC signal and result in abnormal engine start characteristics on the ground or lack of engine response to commanded thrust levels in flight. Although there have been no LOTC events attributed to icing of the P3B sensing system in the field, inspections have identified moisture in this system, which could freeze and corrupt the P3B signal to the FADEC as well. This condition, if not corrected, could result in LOTC due to blockage of the FADEC sense lines, which if it occurs in a critical phase of flight, could result in loss of aircraft control.
Simultaneous LOTC Events
The FAA is especially concerned about the possibility of simultaneous LOTC events on both engines installed on the Boeing 777 series aircraft due to common mode threats, such as certain atmospheric conditions that may result in ice in the Ps3 or P3B pressure sensing system and causing corrupted signals to the FADEC in both engines.
SUMMARY: This amendment adopts a new airworthiness directive (AD) that is applicable to certain General Electric Company GE90 series turbofan engines. This action requires visually inspecting Ps3 and P3B sense lines and full authority digital engine control (FADEC) Ps3 and P3B sensing ports and fittings, cleaning Ps3 and P3B fittings and sensing ports, purging the Ps3 and P3B systems of moisture, and, if necessary, blending of high metal, nicks, burrs, or scratches on Ps3 and P3B fitting threads. This amendment is prompted by seven reports of loss of thrust control due to corruption of the signals to the FADEC caused by water freezing in the Ps3 sensing system. The actions specified in this AD are intended to prevent loss of thrust control due to corruption of the Ps3 and P3B signals to the FADEC which if it occurs in a critical phase of flight, could result in loss of aircraft control.
Interim Action
Both Ps3 and P3B pressure systems incorporate weep holes that allows drainage of water in the lines that may accumulate from condensation or ingested water; however, the field events and the investigation have determined that these design features may not always be effective in eliminating water from these systems. GE is assessing design changes that will prevent water from freezing in these systems and causing corruption of the signals to the FADEC. The requirements of this AD may change based on the ongoing investigation of the root cause and field inspection results, and future rulemaking may be necessary.
I haven't actually covered this area for a long time so i am probably out of date, but a few years ago I spoke to a Lufthansa engineer who had been running mobile phone interference trials and he told me that the millimetric wave landing systems (I still can't recall the proper name and I'm too lazy to look it up) flood the aircraft as it lands and it is on whether the aircraft is using it or not. The avionics systems are shielded against this, and five years of Lufthansa tests had failed to detect any impact from mobile phones even when they were duct taped to the junction boxes.
Other people, admittedly people with bees in their bonnets, said that take-off and landing are almost universally done by computer these days because they are more reliable than humans and do what ATC tells them to. The only times the pilots get their hands on the controls is during the cruise phase, apparently, but I may be wrong/out of date on this one. Engineers now believe that autonomous aircraft with no pilots are perfectly feasible, though this crash may have proved them wrong.
I don't think you can apply specific problems with GE engines to RR!
As for the landing question, it's quite a good one. I'm active on PPruNe, and although the flybois over there can be a bit prickly when self loading freight ask technical questions, I may risk it.
Autonomous planes - yeah, right. This incident certainly goes into the pile of reasons why we want people at the pointy end solving problems.
OK - I've gone and asked. Turns out that it depends on different airlines' operational procedures, but most passenger jet flying is on autopilot from about a thousand feet after takeoff to the end of the approach, although landings are generally manual unless weather conditions preclude them.
Meanwhile, the investigation into BA038 is looking at fuel. Jet aviation fuel crystallises into wax at a low enough temperature, which is not something you want to happen. There was a large air mass of exceptionally low temperature en route that day; another pilot flying a similar route at a similar time said that he's never seen anything like it in 25 years, and everyone in the area had to descend to lower levels to get the fuel temperature up.
What may have happened with the BA flight was that the fuel started to wax up, but that the fuel flow during the cruise was enough to prevent that blocking the feed. In the approach, though, the engines are throttled back to flight idle and take a lot less fuel - and that reduced flow may have been enough for the lines to clog. When the time came for them to spool up to produce more thrust just prior to landing, the fuel supply was heavily restricted and they couldn't operate properly.
If this is what happened, it'll be the first such incident in modern aviation history - and the question will be why the fuel temperature monitoring systems and procedures didn't work.
