Delta Boeing 777 engine suffers 'uncommanded rollback' The US's National Transportation Safety Board is probing an "uncommanded rollback" on a Rolls-Royce Trent 895 engine which affected a Delta Air Lines Boeing 777-200ER on 26 November, Flight International reports. NTSB explains that Delta Flight 18, from Shanghai to Atlanta, experienced loss of thrust on the right-hand engine …
It only takes one small bug that only occurs when a very rare set of conditions are met in the code and you've got a non repeatable and potentially dangerous incident. No computer code is 100% bug free and the flight systems on a 777 will be no exception. Getting rid of obvious bugs is easy - getting rid of these sorts is damned hard. Also AFAIK Boeing don't use separate systems designed by separate teams so any bug in the code will occur in the backup systems too.
"are now examining the Delta aircraft's flight data recorder"
Yep, theres a scuff here and a small dent there where someone installed it with a lump hammer.
No doubt there are some dead thunder-flies in there too, from the summer.
Oh wait, do you mean they're examining the data stored on said device?
In the BA report, the findings indicated the the 777's engine control system properly commanded a thrust increase, so if it was a computer glitch, it would have been in the individual engine's computer. Having both engines have the same bug seems just as implausible. The recent cavitation they found in the fuel lines does suggest it could have been ice crystals. Interesting that it happened with the same Rolls-Royce engine model, though.
as noone ever reads them. They'd rather go on spouting their own ill-informed opinions to all and sundry than actually try a bit of research.
Quote from the AAIB report:
"...the engine control system detected the reduced fuel flows and commanded the Fuel Metering Valves (FMVs) to open fully. The FMVs responded to this command and opened fully but with no appreciable change in the fuel flow to either engine."
In other words:
This was a fuel supply blockage, not a software failure. The FADEC actually sensed the situation and tried to recover it by opening the Fuel Metering Valves fully in order to maintain supply to the engines.
This can easily be corroborated by the following paragraph, describing the damage caused to one of the fuel pumps by the lack of supply pressure:
"The HP pump manufacturer conducted tests on a new pump in an attempt to replicate the cavitation marks seen on the accident flight pumps. The test revealed that running the pump with an abnormally low inlet pressure and a restricted fuel flow of 5,000 pph for 60 seconds gave identical cavitation marks to those seen on the pumps removed from G‑YMMM."
Fuel supply blockage. End of story.
The physical evidence on the prior incident showed fairly conclusively that the problem was ice in the fuel, not a computer or avionics problem. (damage to the fuel pump caused by cavitation, or sucking air was observed)
I would have thought that a directive to remedy this problem would have been issued by now...
I don't know where within the engine system the flight data recorders get their feed of data or signals. But I would have thought they would be recording the actual messages over the databus which feed the engine control systems, so if an engine throttle demand message was transmitted then that would have been recorded, thus, if a software bug present in the engine control system resulted in a reduction in engine power, a reduced power setting message would have been transmitted over the bus, then this would be easily identified.
All you then need to do, is marry up the messages sent from the pilot's throttle levers over the databus and it should be easy to identify whether there was a software bug or not.
All the engines have the same engine control electronics and the same engine control software and share many common inputs, so common failure modes would actually not be that surprising. As another commenter noted, there is no dissimilar redundancy or independent design in this setup. What would be surprising would be common mode problem leading to service failure; that kind of stuff is supposed to be designed out.
That being said, the AAIB report quoted by Richard seems fairly conclusive that the engine control did the right thing (I've read the whole report), and the fuel was the problem. Which is a shame, I quite liked the idea that Broon's spooks were using a high powered cellphone jammer which fried the controllers or some of their inputs. But why let facts spoil a good rumour or rant.
I suppose they could be a heater in to the fuel, prevent the ice forming, but you know what would happen, there would have to have a control system to turn the heater on when at high altitude in freezing conditions for a certain length of time and then turn it off when at lower altitude, so a computer...with a software bug... which would over heat the fuel and kaboom!
300 people dead and they're they'd be trying to track down an almost near impossible problem..back to square 1.
Computers...who needs 'em
I am loth to agree with the NTSB concerning engine fuel problems, however the explanation given for the apparent engine fuel starvation, is certainly more feasible than the puerile comment given by the UK AIB. I would have thought that by now defect monitoring procedures would be in place to obviate any further recurrence of this apparent phenomena.
I am not fully conversant with the B777 but I assume that the fuel heaters are sufficient for the job and I am sure that many sectors of high duration have been flown worldwide with this aircraft.
Let us hope that a workable solution to this problem will be forthcoming soon, for the sake of Air Safety.
Or you could follow my link, 2nd from top, and find a possible solution which doesn't involved heaters or software. And it's cheap. In fact here's the text in full so no-one has to follow a link.
"
If the Boeing 777 was indeed brought down at Heathrow by ice in its fuel (13 September, p 23), it would not be the first. One night in 1948 or 1949 four Royal Navy Attacker FB2 aircraft "fell out of the sky" at Royal Naval Air Station Milltown in Aberdeenshire, UK. Each had almost total loss of power in its single Rolls-Royce Nene 102 engine. In the morning we could find nothing wrong, apart from water in their low-pressure fuel filters when they were drained - about the same amount in each. They all started without trouble and were ground-run successfully. We concluded that the filters had iced up to the extent that the fuel flow had been severely impeded, but that overnight the ice had melted and the problem vanished.
Careful investigation seemed to exonerate home base fuelling arrangements from being the source of the water, so we had to look elsewhere. Much of the fuel in the Attacker was carried in an exposed, virtually uninsulated, ventral tank. This would suffer severe cooling at altitude.
Aviation fuel at room temperature contains dissolved water. We thought this might have come out of solution and formed ice on the filters. From then on it was standard procedure to add a small amount of methanol as antifreeze when refuelling the ventral tank. As far as I know, there were no further problems of this kind.
Since no one was injured and no damage done the press evinced no interest. Possibly by now there are not a lot of us left who remember the incident, but Rolls-Royce may still have the details in their archives.
"
The Register 
Biting the hand that feeds IT
Copyright. All rights reserved © 1998–2024
