Another 777 uncommanded engine rollback
Thread Starter
Join Date: Jan 2005
Location: France
Posts: 2,315
Likes: 0
Received 0 Likes
on
0 Posts
Another 777 uncommanded engine rollback
NTSB e-mail, December 18, 2008
NTSB INVESTIGATING LOSS OF ENGINE POWER ON DELTA AIR LINES BOEING 777
The National Transportation Safety Board is investigating an
incident in which a Delta Air Lines Boeing 777 experienced
an uncommanded engine rollback in the cruise phase of an
intercontinental flight.
On November 26, 2008, at about 12:30 pm MST, in the vicinity
of Great Falls, Montana, a 777-200ER (N862DA), operated by
Delta Air Lines as Flight 18, en route from Shanghai to
Atlanta, experienced an uncommanded rollback of the right
(number 2) Rolls-Royce Trent 895 engine while at 39,000 feet
in the cruise phase of flight. The crew executed applicable
flight manual procedures and descended to 31,000 feet. The
engine recovered and responded normally thereafter. The
flight continued to Atlanta where it landed without further
incident. None of the crew of 15 or 232 passengers was
injured.
Flight data recorders and other applicable data and
components were retrieved from the airplane for testing and
evaluation. Both of the pilots have been interviewed.
This event is preceded by another airline's 777 equipped
with Rolls-Royce Trent 895 engines, which experienced an
uncommanded dual engine rollback while on final approach to
London's Heathrow International Airport on January 17, 2008,
crashing short of the runway on airport property. The
United Kingdom's Air Accidents Investigation Branch (AAIB)
is investigating that accident.
NTSB Senior Air Safety Investigator Bill English, who is
serving as the U.S. Accredited Representative in the
Heathrow accident investigation, is the Investigator in
Charge of the Delta incident.
The AAIB, which has assigned an Accredited Representative to
the Delta incident, is working closely with the NTSB to
determine if there are issues common to both events.
The National Transportation Safety Board is investigating an
incident in which a Delta Air Lines Boeing 777 experienced
an uncommanded engine rollback in the cruise phase of an
intercontinental flight.
On November 26, 2008, at about 12:30 pm MST, in the vicinity
of Great Falls, Montana, a 777-200ER (N862DA), operated by
Delta Air Lines as Flight 18, en route from Shanghai to
Atlanta, experienced an uncommanded rollback of the right
(number 2) Rolls-Royce Trent 895 engine while at 39,000 feet
in the cruise phase of flight. The crew executed applicable
flight manual procedures and descended to 31,000 feet. The
engine recovered and responded normally thereafter. The
flight continued to Atlanta where it landed without further
incident. None of the crew of 15 or 232 passengers was
injured.
Flight data recorders and other applicable data and
components were retrieved from the airplane for testing and
evaluation. Both of the pilots have been interviewed.
This event is preceded by another airline's 777 equipped
with Rolls-Royce Trent 895 engines, which experienced an
uncommanded dual engine rollback while on final approach to
London's Heathrow International Airport on January 17, 2008,
crashing short of the runway on airport property. The
United Kingdom's Air Accidents Investigation Branch (AAIB)
is investigating that accident.
NTSB Senior Air Safety Investigator Bill English, who is
serving as the U.S. Accredited Representative in the
Heathrow accident investigation, is the Investigator in
Charge of the Delta incident.
The AAIB, which has assigned an Accredited Representative to
the Delta incident, is working closely with the NTSB to
determine if there are issues common to both events.
Join Date: Mar 2006
Location: Choroni, sometimes
Posts: 1,974
Likes: 0
Received 0 Likes
on
0 Posts
Investigator Bill English, who is
serving as the U.S. Accredited Representative in the
Heathrow accident investigation,
serving as the U.S. Accredited Representative in the
Heathrow accident investigation,
Join Date: Apr 2002
Location: FUBAR
Posts: 3,348
Likes: 0
Received 0 Likes
on
0 Posts
Indeed, a slightly disturbing incident for sure. I wonder if a common cause or just some random electrical spike. Perhaps all the research into ice etc will prove to have been for nought and the cause for both found (or not ) to have been some transitory electrical woopsie that nobody forecast and the BITE stuff doesn't recognise.
Damned funny stuff electricity, black magic.
Damned funny stuff electricity, black magic.
I was amazed there was no response from Ppruners.
There are more common reasons for rollback not related to ice in the fuel.
The BA038 thread evidenced unique response of the fuel control not associated with more common causes of rollback.
I'm not saying that the fuel control is causal, only that's a indicator of where to look.
Join Date: Sep 2005
Location: uk
Posts: 267
Likes: 0
Received 0 Likes
on
0 Posts
sems to dispell the ice in fuel explanation of the beijing incident.becoming a bit concerning if software and electrical systems are less than 100% reliable.considering the number of triples in service,i suppose no need to cause too much concern.
Don’t discount the weather as a cause:- The Ice Crystal Weather Threat to Engines.
becoming a bit concerning if software and electrical systems are less than 100% reliable.
Join Date: Nov 2004
Location: Nowhere
Posts: 60
Likes: 0
Received 0 Likes
on
0 Posts
"becoming a bit concerning if software and electrical systems are less than 100% reliable."
I doubt any are, but they can still be more reliable then a large number of mechanical moving rods.
I doubt any are, but they can still be more reliable then a large number of mechanical moving rods.
a bit concerning if software and electrical systems are less than 100% reliable
Join Date: Sep 2007
Location: Paris, France
Posts: 350
Likes: 0
Received 0 Likes
on
0 Posts
The problem is that software isn't verified as thoroughly as hardware.
Verification and certification processes were designed for mechanical systems, not ones and zeroes. Digital systems, including anything controlled by software, have failure modes that are dramatically different from those of mechanical systems, and catastrophic failures for unforeseen circumstances are the rule rather than the exception. Systems that mix software and hardware are still tested and verified as if everything were hardware, and that lets lots of potential bugs slip through the cracks.
Combine this with the extremely poor record of reliability for software engineering as a whole, and software becomes a considerable cause of concern.
I know that aviation tries harder to verify software than most other industries do, but manifestly it isn't trying hard enough, if events like this are happening. Which type of mechanical failure would roll an engine back in cruise? Weird, random, off-the-wall incidents like that have all the hallmarks of software bugs.
My specialty is information technology; aviation is my hobby. Decades of experience with IT have taught me that the more software there is in a safety-of-life system, the less it should be trusted. In an ideal world, software would be completely safe, but we are still in the Dark Ages when it comes to designing reliable software. To paraphrase a famous saying, whenever I hear "computer," I arm the pyrotechnics on my ejection seat.
Verification and certification processes were designed for mechanical systems, not ones and zeroes. Digital systems, including anything controlled by software, have failure modes that are dramatically different from those of mechanical systems, and catastrophic failures for unforeseen circumstances are the rule rather than the exception. Systems that mix software and hardware are still tested and verified as if everything were hardware, and that lets lots of potential bugs slip through the cracks.
Combine this with the extremely poor record of reliability for software engineering as a whole, and software becomes a considerable cause of concern.
I know that aviation tries harder to verify software than most other industries do, but manifestly it isn't trying hard enough, if events like this are happening. Which type of mechanical failure would roll an engine back in cruise? Weird, random, off-the-wall incidents like that have all the hallmarks of software bugs.
My specialty is information technology; aviation is my hobby. Decades of experience with IT have taught me that the more software there is in a safety-of-life system, the less it should be trusted. In an ideal world, software would be completely safe, but we are still in the Dark Ages when it comes to designing reliable software. To paraphrase a famous saying, whenever I hear "computer," I arm the pyrotechnics on my ejection seat.
Thread Starter
Join Date: Jan 2005
Location: France
Posts: 2,315
Likes: 0
Received 0 Likes
on
0 Posts
AnthonyGA,
Yes, no and maybe.
In the analog days, we had TWO physically separate and distinct channels, one monitoring the other. If they disagreed, they disconnected. The probability of the same op-amp or same resistor failing identically at the same time was infinitesimal.
When 'digital' came along, I had qualms as well, but dissimilar processors and dissimilar software between 'command' and 'monitor' mostly were supposed to have solved their 'identical failure' problems.
Some of those concepts now seem to have gone out of the window, judging from some recent posts elsewhere.
So far, we have no idea whether this incident is again ice or something similarly mechanical, or indeed "IT" related, so let's give it a break until we have more data?
Cabin crew spilled some Coke on the pedestal during the last flight and it now soaked through? Weirder things have happened.
I sincerely hope the QAR/FDR data will tell a useful story.
CJ
Yes, no and maybe.
In the analog days, we had TWO physically separate and distinct channels, one monitoring the other. If they disagreed, they disconnected. The probability of the same op-amp or same resistor failing identically at the same time was infinitesimal.
When 'digital' came along, I had qualms as well, but dissimilar processors and dissimilar software between 'command' and 'monitor' mostly were supposed to have solved their 'identical failure' problems.
Some of those concepts now seem to have gone out of the window, judging from some recent posts elsewhere.
So far, we have no idea whether this incident is again ice or something similarly mechanical, or indeed "IT" related, so let's give it a break until we have more data?
Cabin crew spilled some Coke on the pedestal during the last flight and it now soaked through? Weirder things have happened.
I sincerely hope the QAR/FDR data will tell a useful story.
CJ
Join Date: Jun 2005
Location: Cheshire, UK
Posts: 160
Likes: 0
Received 0 Likes
on
0 Posts
Cabin crew spilled some Coke on the pedestal during the last flight and it now soaked through? Weirder things have happened.
Everywhere I have ever been in aviation, everyone who was allowed on the flight deck was taught to pass drinks to pilots around the outside and never, ever on the inside over the central pedestal.
Fluids and automatics do not mix and exciting and expensve things happen when they come together.
Fluids and automatics do not mix and exciting and expensve things happen when they come together.