Do I understand correctly that SF is an unusually difficult airport to land at?
Other than the final approach is over water I don’t see it as being difficult. A will of course defer to any pilot that has actually landed there.
If this plane was actually close to stall speed then it pulling up hard would have made it worse. They would have been better off keeping it level to recover enough speed to go around.
Another thing to consider is the cultural aspects of the flight. From my understanding in Asian countries there is still a mentality of “seniority rules” in the decision making process that use to be prevalent in all of aviation. This was suppose to have been addressed after the Tenerife Airport double 747 crash. For those not familiar with it there were junior officers who questioned the PIC’s decision but did not press it.
The view in the cockpit is the same for any normal approach. With the plane trimmed out on final the runway remains in the same position visually so it should have looked wrong from the get-go if they were low to begin with.
Rick is right. I have landed on that runway more than a few times and there is nothing abnormal about it at all.
Landing over water can add some visual illusions but on a clear day they would be minor at their worst. Now, if you were doing the approach at night and the runway had an extreme up slope then you can get into some funky illusions but that is not the case here.
Landing on a runway like this the pilot should be aware of the seawall and the water but it is pretty common in the US. LaGuardia in NYC, Reagan National in DC and Boston come to mind as similar runways off the top of my head. At least SFO has nice long runways unlike LGA or DCA.
To add to what JCFindley said, there are runways and approaches that require special training because they don’t allow for a standard approach. This isn’t one of them.
This is a serious (not snarky) question prompted by this headline and article: “Asiana Crash A Point Of National Shame For Koreans”: will some high up person in the company be expected to commit suicide over this? (As sometimes happens in Japan?)
tv news showed photos of passengers who had evacuated the aircraft taking their carry on luggage.
Yes that is not uncommon in evacuations. Unfortunately if the crew had try to force the passengers to leave their baggage behind it would only have slowed the evacuation down more. I’ve seen someone suggest the overhead lockers should be locked for take off and landing. That would stop people trying to stand up and get their bags before the seatbelt sign was off on a normal flight as well.
I’ll be interested to see whether the FDR shows pilot-induced oscillations on final approach. Without doing a whole lot of research (and obviously without the NTSB’s final report) all I have to go on is a hunch, but it certainly wouldn’t surprise me, PAPI or no.
“Pilot-induced ocillations”?
Minor over corrections to flight path deviations resulting in the aircraft wobbling around the desired flight path. Sometimes aeroplanes fly best when the pilot doesn’t do anything silly, like move the controls ;).
Ah, thanks. I’ve done something similar when trying to keep a sailboat on course (I’ve only been at the helm maybe 20 times in my life).
Could this phrase refer to wobbles in either the vertical or horizontal planes?
Yes. Mainly I’ve seen it (and done it myself) in the form of wing waggling on late approach, particularly with full landing flap. 99% of the time it’s no big deal, it’s typically just due to a lack of finesse and/or experience on the aircraft.
Brings up an interesting point. Disclaimer: I’m a mere private pilot–Cessna driver.
I think the layperson has a hard time putting himself in a pilot’s place because aircraft are so unlike cars, which nearly everyone has experience with.
Barring mechanical failure, airplanes naturally want to fly themselves. In other words, once the pilot sets a certain speed and selects a certain attitude (nose up 5 degrees, for example) and has trimmed the airplane, the plane itself wants to settle into that flight regime and stay there. Even without the autopilot engaged, the pilot flying can take his hands and feet off the controls and the plane will find its equilibrium for that specific speed/attitude. If the controls are manipulated, then the pilot lets go, generally the plane will “find” that equilibrium again.
So in a final approach, if you’re doing everything “right,” once you’ve lined up laterally with the runway (either visually or by establishing the ILS localizer) and established the sink rate (i.e. you’re on the glideslope) then in a stable approach, there’s not a lot the pilot has to “do” to land the airplane. If you end up a little high, then you reduce power and re-intercept the glideslope. If you’re a little low, you add power and do the same.
Now, especially for a beginner–or, I’m assuming, when you’re new in-type–you just psychologically have a tendency to want to “overfly” the airplane instead of letting it settle into its own thing. That’s one of the reasons you see the wing-wagging or pitch movements on final (especially if you’re hanging around an airport near a flight school.) I can’t speak for other pilots, but for me, overflying was probably the most difficult habit to break, except for having to break yourself of the habit of trusting your inner ear when flying blind.
PIO’s are a pretty common (and not usually ‘dangerous’) result of this overflying tendency.
Another factor in PIOs is that large aircraft can be surprisingly light in the controls. The feel is normally artificial and a small control movement can have a large effect. Compounding this is that at the relatively high approach speeds of a passenger jet, a small change in pitch causes a large deviation in the flight path. You have to be quite delicate with the controls to fly them nicely.
There is no description of porpoising from the passengers or outside observers. From what I’ve heard the approach was set up for a certain speed and for whatever reason that didn’t occur and it was caught too late. Of course we don’t have the facts beyond the airspeed was incorrect and this put them behind the power curve. The wrong information could have been programmed into the computer or some other failure led the plane to adjust the engines incorrectly or they could have been hit with some weird-ass change in wind speed. The PIC is ultimately responsible for situational awareness and control of the aircraft.
It does seem to me, from the hindsight perspective of a low time SEL pilot, that the correct response to the situation was NOT to radically pull up. They were close to stall that even with FADEC controlled engines it’s not going to spool up fast enough for a standard go-around. They would have had the full length of the runway to build up speed. They only needed to gain enough altitude for the landing gear to clear the sea wall. Ground effect would allow for that maneuver. If necessary they could have deliberately taken out the approach lights. Wouldn’t be the first time a pilot did that.
The news this morning is that the autothrust which automatically controls speed was armed but not turned on, and that the pilots were concentrating on lining up with the runway and didn’t notice that their airspeed had dropped until it was too late. They testified that the autothtust was on - so it seems they must have thought it was, and paid no attention to their speed.
And sometimes not at all minor.
To be blunt, I had a love/hate relationship with auto throttles and generally if something seemed hinky, that was the first thing I clicked off.
The B717 was the only aircraft I flew that had good auto-throttles. They would keep your airspeed nailed if you were stabilized but if you were even a little behind the jet, they tended to be a bit extreme in their corrections and they can’t predict what you are doing if you are hand flying the jet.
For instance, if you are above glideslope and have a high descent rate you should be pushing the throttles up as you approach glideslope while or before you pull back on the yolk to intercept it. The autothrottle only adjusts for what the airplane is doing now though as opposed to what you are going to do with it.
Then, there were times when I thought they were on but they weren’t. Yes, I have done that one myself a time or three BUT it should be inherently obvious that the airplane is not doing what you want it to.
I may be a bit old school in this but if the airplane wasn’t doing what I thought I programmed the automation to make it do, I clicked off all the automation and hand flew it.
My favorite quote on the subject is from an instructor in Air Force Pilot Training from long ago.
“Dude, just be a pilot.”