Well, the real point is was this a "stupid pilot" error or an error
the pilot made when he though he was doing something else, or
alternatively a training issue where the pilot did something he should
not have when he thought he was doing something else, and what he did
was what he was taught to do.
One does NOT land one of these things like one does a single engine
Cesna on a clear day with no wind, I suspect the only time they are
landed without autothrottle control is when something is wrong with
the aircraft. It's foolish to manually control the airspeed when the
autospeed does a better job anyway.
We will see, the NTSB generally does a superb job of accident
investigation, and what exactly happened will get ferreted out
eventually. Probably going to take some time, though.
This is similar to the AA DC-10 crash in Chicago back in the mid 70's
where the engine came off just after liftoff. The crew failed to
correct for the loss of lift on the affected wing due to leading edge
slat retraction upon hydraulic failure and the aircraft rolled over
and crashed. It was very easy to compensate for, all the crew had to
do was retract the slat on the other side to balance the lift, so you
could call that pilot error. However, since the power for the mis-
match slat extension warning was supplied only by the engine that
ripped off, there was no warning and the plane rolled over too far to
recover before the crew could fix it. They DID cotton on the the
problem, but only after it was too late to save the plane.
Was that pilot error? No, every single DC-10 pilot who flew the
"accident" on a simulator crashed the plane, every time. If the slat
mis-match warning was enabled, every single one flew it out with no
issues and "landed" safely.
Pilot error? Not really.
If the Asiana crew thought the autothrottle was set and working, as it
always was on landing a 777, and it did NOT control the speed as the
aircraft approached the set landing speed AND there was no audible or
visual warning that the speed was not being controlled by the
autothrottle, is that pilot error or a training/equipment problem?
There is, from what I've read, at least one flight mode where the
autothrottle is deactivated AND the speed warning/autothrottle failure
is ALSO deactivated (look up "FLCH trap"on the 777).
Remember, SFO ATC is well known for not giving final approach
clearance until the landing aircraft is past the inital flight slope
and too high, requiring some fancy flying to get the plane down to
glide slope and into stable approach conditions. The ALPA has been
complaining for decades, this is not a new problem.
So think this scenario: ATC doesn't give landing clearance until the
ASIANA flight is past the initial point on the glide slope. This is
against company rules, since it requires an un-stabilized approach,
but it's routine at SFO. It's also not a big problem, but to get down
to glide slope, it's necessary to idle way back on the engines and
descend a bit faster than normal. This is a computer operated plane,
you don't manually handle the throttles much, you select the speed you
want and the rate of descent. However, since you are over the glide
slope, you have to use some flight mode on the computer that lets you
do so, and in this case, if the crew selected FLCH (flight level
change) and did NOT realized that doing so deactivates the auto
throttle AND the auto throttle "failed to engage -- speed too low"
warning, we are in a situation where the crew expects the autothrottle
to control the speed and is NOT watching the rate of speed loss very
carefully, they are watching to make sure they achieve glide slope and
don't have to go around to avoid an over-run by landing too fast too
far down the runway.
During the last few seconds of the approach, the engines don't come
up, speed starts to drop below set landing speed, crew has to figure
out why (remember, they always use auto throttle speed control), and
oops, crash. By the time the failure of the engine power to come up
when expected gets processed by the crew and they ram the throttles
wide open, it's about 10 seconds too late to avoid the crash.
Throttle application on a 777 is timing critical, that bus descends
fast and takes a lot of thrust to stop a descent. By the time the
speed drops below selected landing speed you are at least 5, probably
10, seconds LATE on throttle application and there is absolutely
nothing that can be done to fix it, you are going to stall.
Obviously, the crew expected the speed to be automatically controlled,
and it was a very nasty surprise when it wasn't.
Now, where is the error? Is it "stupid pilots" who recklessly fail to
set the autothrottles? Is it improper training -- never use FLCH mode
to descend to glide slope (did anyone make sure the pilots knew FLCH
would leave them with NO autothrottle and NO wanings)? Did SFO ATC
give them a grossly inappropriate approach, leaving them to decide
after 11 hours in the air to demand a go-around at a busy airport,
burning lots of extra fuel and causing delays or to land without a
stabilized approach?
Not a simple issue and just saying "stupid pilot can't fly" isn't the
answer. Training might very well be part of the issue (it usually is
from what I've read), but a very large portion of safety work,
including flying and landing aircraft, is to make sure people CAN"T
make mistakes that crash planes, blow up plants, cut peoples body
parts off, etc. It should not be possible to select a computer
control mode during final approach that disables both autothrottle AND
speed failure/autothrottle not activated warnings, as doing so will
very often cause the plane to crash even if the pilots are monitoring
the speed. PIlot error or bad software/instrumentation design?
Peter
_______________________________________
http://www.okiebenz.com
For new and used parts go to www.okiebenz.com
To search list archives http://www.okiebenz.com/archive/
To Unsubscribe or change delivery options go to:
http://mail.okiebenz.com/mailman/listinfo/mercedes_okiebenz.com
