alt Hacker NewsSo many mistakes. A notable part is this:
>> CVR: 00:15 A pilot called for a go-around.
>> FDR: 00:15 Brakes were released at a groundspeed of 120 knots and throttles were increased to a value consistent with go-around thrust.
>> FDR: 00:11 The throttles were reduced to idle power and the brakes were re-applied. The groundspeed was 117 knots. The throttles remained at idle, and brakes remained applied until the end of the recording.
So they decided to go-around (i.e. do not land and climb immediately, going full thrust), then 4 seconds later changed their mind and went back to power idle and brakes (unclear how speed didn't increase).
While they were still late in calling a go-around (they should have done it before being anywhere close to touching down due to their unstable approach), guessing those 4 seconds contributed to the disaster by a lot.
Replies
> While they were still late in calling a go-around (they should have done it before being anywhere close to touching down due to their unstable approach)
Barring a massive tailwind, they had enough groundspeed to go around at touchdown (158 kts).
I think we'll find throttling to G-A and then aborting the G-A caused the crash. If they'd committed to landing they at least wouldn't have overshot with that much energy. If they'd decided to G-A even at 00:15, they may have been well within margin to take back off.
> (unclear how speed didn't increase)
Large high-bypass turbofan engines take several seconds to spool up from idle (as they would be on landing) to full thrust. The 4 seconds that the thrust levers spent in the full thrust position in this accident was not long enough.
The go-around decision was far too late. Remember the engines need significant time to spool up. The decision to cancel the go-around was correct (conditional on the preceding series of terrible decisions).
mind you those numbers are seconds until the cvr stopped recording. they were a few hundred feet from the end of the runway when they initiated the go around and like the other poster said the turbines do take time to spool up from idle. they were so far out of normal spec for landing that it was absolutely pilot fault but not at the far end of the runway. they were not stabilized on approach way further back about 5 min or so. (if you're not stabilized on approach you are required to go around). they even called for flaps 30 halfway down the runway which is also not allowed for a stabilized approach.
Spool up time is 8-10s on high-bypass turbofan engines. They tried to go around maybe 1000ft from the end of the runway? It was far far too late to even try it at that point.
Sadly if they had made the call earlier they were going so fast they were basically at rotation speed when they touched down so could have just pulled back and gotten into the air.
They landed flat (and the left wing refused to come down) because they tried to force the plane onto the runway. That just doesn't work. The wings are creating too much lift when you're that fast.
Disclosure: Pilot but not ATP or Jet rated.
> unclear how speed didn't increase
How can you comment as if you are knowledgeable enough about aviation to speak on this with authority, but you aren't aware that turbofan engines don't have instant throttle response and need time (up to 8 seconds) to spool up? I think a little more restraint is warranted in something like this if you don't know something that basic.
You can actually do the math to work backwards and approximate how much thrust the engines were producing during that time, which was about 15,000 lbf (vs ~120,000 lbf max thrust for a 767-300er) which reduced how quickly they were decelerating by... Maybe a few % or so. But they didn't start accelerating yet. They would have in another second or two, however.
I'll agree that they should have gone around far earlier; getting non-spurious GPWS callouts on final because you're trying to dive and find the glideslope is... Crazy, and configuring landing flaps over the numbers is also crazy, and so is landing 10-20kts faster than the highest possible Vref value for the type, and everything else. But that's fairly obvious...