Korean 777 Down in SFO

The captain of the GLMI glider would undoubtedly have landed this aircraft without a problem. His earlier training made a VFR sunny day landing, engines out , not only a possibility but a reality. ( including a nice slip at the end) These fellows had engines that were producing yet the whole cockpit full of pilots could not get it safely on the ground. If it was a cultural problem, dependence on auto throttle ( which was off) etc. etc. , whatever....it was a massive f.... Up.
 
Last edited:
I read the whole CVR transcript, they didn't sound stressed out to me except maybe unrelated job/business stress. They made their altitudes and speeds just fine, they just didn't clue into the visual perspective cues on short final and got too slow.

Their airspeeds were NOT fine. They were 30 knots below VRef on final.
 
Their airspeeds were NOT fine. They were 30 knots below VRef on final.

I was referring to them hitting their target speeds and altitudes at their crossing points. Someone said they were slammed in, and it it didn't appear that way, they got it slowed down just fine.
 
I was referring to them hitting their target speeds and altitudes at their crossing points. Someone said they were slammed in, and it it didn't appear that way, they got it slowed down just fine.

They got it slowed down to just above a stall. (I'd add LOL, except that it really isn't funny.)

No way YOU'D let that happen. On final, you're looking at the airspeed slowing through 110 kts--I think you'd at least shout at someone! :yes:
 
They got it slowed down to just above a stall. (I'd add LOL, except that it really isn't funny.)

No way YOU'D let that happen. On final, you're looking at the airspeed slowing through 110 kts--I think you'd at least shout at someone! :yes:

Nope, as soon as I saw the threshold sliding up the window I would have advanced the power levers. I don't shout unless it's to be heard over conditions that warrant it, I have never found it to be productive.
 
Nope, as soon as I saw the threshold sliding up the window I would have advanced the power levers. I don't shout unless it's to be heard over conditions that warrant it, I have never found it to be productive.

That's it. Yet we had, what--five pilots?--all sitting dumb and happy while their perfectly functioning ship flew into the ground without a skipper.
 
That's it. Yet we had, what--five pilots?--all sitting dumb and happy while their perfectly functioning ship flew into the ground without a skipper.

Yeah, from the CVR transcript, nobody even noticed they were screwed until the plane started chiming 'sink rate', and even then it seems to have taken a bit before anyone reacted.
 
Yeah, from the CVR transcript, nobody even noticed they were screwed until the plane started chiming 'sink rate', and even then it seems to have taken a bit before anyone reacted.

Which is what totally kills me...watching the video, the extreme pitch angle alone should have given them a clue that something wasn't right.
 
PF was a passenger. I guess they all were.
 
It seems that everyone wants to actually "hands on" fly the airplane. Those days are over. Why use a computer in bad weather and not in good weather? If it works best when the pilot cannot visually see, why isn't that the best way on a clear day? Or am I missing something.

In 1955 I worked at Curtis Wright and was in the back of the cockpit of the first simulator that used an actual cockpit, during acceptance tests by the government. C118, DC6.
 
Visual approach stressful in perfect weather, only to a button pusher.

Oh Noes! Not the dreaded CAVU VFR approach :hairraise:

Reports I've read say although the ILS was out of service the PAPI was functional. Why not follow the PAPI?
 
It seems that everyone wants to actually "hands on" fly the airplane. Those days are over. Why use a computer in bad weather and not in good weather? If it works best when the pilot cannot visually see, why isn't that the best way on a clear day? Or am I missing something.

In 1955 I worked at Curtis Wright and was in the back of the cockpit of the first simulator that used an actual cockpit, during acceptance tests by the government. C118, DC6.

I would say you are missing something. In fact, I would say you are missing a lot of different somethings. I guess this begs a long spiel about automation and human factors. That stuff can be looked up in da googles. In this particular case, the human set the A/T such that it was 'armed' but not 'enabled' and the human possibly thought it was both armed and enabled(I think the transcript uses the word 'hold' mode). So, the A/T did exactly as the designers of the automation system said it was supposed to do, it sat there, and did not advance the throttles as the plane slowed below the target speed. it worked flawlessly right up to the point where the human had to take over and resolve the discrepancy.

So the computer that manages the A/T, and in other modes provides a measure of protection was doing exactly what it was told to do, maintain its position and not move. Good computer, you followed the instruction from the human to perfection. The human waited, and waited, and waited for the A/T to do its thing, and the A/T computer was sitting there waiting for the human to do its thing, and viola!

Perhaps we can build more heuristics into flight computers: "Ah, I see that you set me to 'hold' mode when you really wanted me to engage 'active' mode and provide speed control as you are on final, and want me to manage the thrust. I see, so I will override the 'hold' mode that you set, and put myself in active mode, thus maintaining target speed(1.3 Vso) for ever, and ever, and ever, and ever. Oops, you seem to have moved the levers forward to TOGA, but you didn't deselect active mode, so I will ignore your thrust request to 100% rated power because you didn't disengage the A/T, so you don't get any thrust, because I decide when and how to apply thrust." Yeah, that's better.
 
Why was it disconnected in the first place? Why did the pilot take control at any time?
 
Just reply to the question: Why did the pilot take control at any time? Period.
 
Why was it disconnected in the first place? Why did the pilot take control at any time?

Notam ILS glide path was out of service for maint or upgrade at the time. No ILS glide path, no vertical guidance - hence hand fly it(or more accurately, hand throttle it).
 
Why was it disconnected in the first place? Why did the pilot take control at any time?

The better question would be, "Why did the four (4) trained and (putatively) competent pilots each and all fail to monitor the progress of the aircraft in its flight, and take immediate and appropriate action when it deviated from a proper flight profile.

Whether the plane is being flown by the autopilot in IMC to a CAT III autoland, or by a pilot in flawless weather, it is the responsibility of the flight crew to maintain contemporaneous knowledge of the aircraft's status, and take whatever action is required to preserve the safety and security of the aircraft and its occupants.

IOW: Fly the damned plane!
 
No vertical guidance to feed to the computers.

It amazes me that it doesn't have an R-Nav system with V-Nav capabilities, I'm not sure that it doesn't since I recall something to those sorts mentioned in the brief for the missed procedure.
 
It amazes me that it doesn't have an R-Nav system with V-Nav capabilities, I'm not sure that it doesn't since I recall something to those sorts mentioned in the brief for the missed procedure.

Of course it has VNAV. But, the tool behind the yoke has to have it selected with a target in the window.
 
The trouble here is that they are not hiring pilots, who have learned how to FLY an airplane. How many flew Beaver floatplanes every summer for 10 years before they became "airline" pilots?
Maybe they have some beaver pilots of staff, hard to say--if that make you a good pilot and all. Agreed that is does seem like some of these carriers staffing is filled with not so skilled pilots.
Button pushers................
Visual approach stressful in perfect weather, only to a button pusher.
Maybe not stressful, but I do remember hearing when I was in INDOC for my first airline job that the hardest thing you will have to learn when you get to the line is how to do a visual approach. You spend many many hours flying the sim and doing approaches, systems, procedures etc.. You know really well how to configure the a/c to get the desired results when you know the procedure in front of you.The sim is not that great at replicating a visual condition. You really never know where you are going to be in reference to the field when they could give you a "cleared for the visual!' SO figuring out what the airplane wants. It was different to do a visual because some of the guidance cues that you are used to 95% of the time is not there. Anyway, not making any excuses for this because really its was not that hard, just different. They screwed up big time.

"Very stressful?" B.S.

Weight, airspeed, aim point. Yes, it might be harder in a heavy, but they were 30 knots underspeed. That has nothing to do with something "visual."
Sure it does..they had to look outside! :rolleyes: Autothrottles will take care of the rest. :nonod:

Which is what totally kills me...watching the video, the extreme pitch angle alone should have given them a clue that something wasn't right.
Heck it seems like they don't know what the normal deck angle is by referencing the horizon.

Why was it disconnected in the first place? Why did the pilot take control at any time?
Because they can.

Notam ILS glide path was out of service for maint or upgrade at the time. No ILS glide path, no vertical guidance - hence hand fly it(or more accurately, hand throttle it).
I bet there is a way to set up the FMS to vertical guidance for runways with no ILS, but I don't think they were that sharp.

It amazes me that it doesn't have an R-Nav system with V-Nav capabilities, I'm not sure that it doesn't since I recall something to those sorts mentioned in the brief for the missed procedure.
I think you are right.
 
I was under the impression that Sullenberger landed in the Hudson River by hand flying the airplane. Was I wrong in assuming this?
 
I was under the impression that Sullenberger landed in the Hudson River by hand flying the airplane. Was I wrong in assuming this?

Well it was an Airbus so Sullenberger only got to vote on aircraft control. As I understand it Air France has specified that the Airbus be stalled and the stall held until landing as a normal emergency procedure. It's a good thing that Sully's vote counted...
 
The best narrative on what happened in those major investigations tends to be in the factual report of the operations group chairman. It is the technical experts speaking before the political appointees get a chance to put their spin on it.
 

Attachments

  • Operations_summary_Asiana.pdf
    550.5 KB · Views: 23
And here if anyone wants to read the 4 different stories from the cockpit on what happened.
 

Attachments

  • Interviews_Asiana.pdf
    1.1 MB · Views: 15
I would bet the state dept. would also be involved saying things like " let's not be too critical. We do a lot of business with South Korea . Lots of money involved here. " etc. etc.
 
The best narrative on what happened in those major investigations tends to be in the factual report of the operations group chairman. It is the technical experts speaking before the political appointees get a chance to put their spin on it.

Yep, good read there, saw the PAPI go from 2 reds to 3 reds and pulled the nose up because if it went to 4 he'd be embarrassed.
 
Yep, good read there, saw the PAPI go from 2 reds to 3 reds and pulled the nose up because if it went to 4 he'd be embarrassed.
Excellent airmanship, that's just what I was taught to do (not).

Pitch for airspeed, power for altitude...(now all the peanut gallery will sound off)
 
Excellent airmanship, that's just what I was taught to do (not).

Pitch for airspeed, power for altitude...(now all the peanut gallery will sound off)

Well, with autothrottles engaged, you do pitch up and the AT maintains the speed. Problem is, with the combination of buttonpushes used at 1500ft, they had inadvertently disabled the AT.

Go to page 26/27 of the interviews where the PF explains that by culture and company policy, the decision for a go-around has to come from the more senior pilot, in this case the IP sitting in the right seat.
 
Well it was an Airbus so Sullenberger only got to vote on aircraft control. As I understand it Air France has specified that the Airbus be stalled and the stall held until landing as a normal emergency procedure. It's a good thing that Sully's vote counted...

More myths.....:rolleyes2:
 
Never occurred to him to add power???

From the sounds of it, he thought the auto throttle would take care of it, he didn't realize the auto throttle was disengaged. I'm not certain why the IP was doing things without calling them out, or waiting for the instructions to do them.
 
They got slam dunked - happens to everyone who flies either IFR or VFR with FF in busy airspace.

They created a 1500fpm rate of descent to make up for it but still were able to slow down to flap speeds given the power needed to keep 425,000# aloft.

They blew through the glideslope with that rate of descent and somehow did not notice it. Then they noticed they were low, but not slow, and raised the nose, screwing up their approach even more. As we all know, pull up go up, push down go down, pull up more- go down.

The approach was not stable, it was not on speed, on heading and on glideslope. The pilots had 10 hours to program the FMS with the virtual glideslope for R28L . . .

oh well.
 
Which is what totally kills me...watching the video, the extreme pitch angle alone should have given them a clue that something wasn't right.
Extreme pitch angle and extreme force - it was said that in the last phase they were pulling close to 80 lbs on the yoke, even a blind pilot could tell something wasn't exactly right :hairraise:
 
They got slam dunked - happens to everyone who flies either IFR or VFR with FF in busy airspace.

They created a 1500fpm rate of descent to make up for it but still were able to slow down to flap speeds given the power needed to keep 425,000# aloft.

They blew through the glideslope with that rate of descent and somehow did not notice it. Then they noticed they were low, but not slow, and raised the nose, screwing up their approach even more. As we all know, pull up go up, push down go down, pull up more- go down.

The approach was not stable, it was not on speed, on heading and on glideslope. The pilots had 10 hours to program the FMS with the virtual glideslope for R28L . . .

oh well.

The glideslope was out of service.
 
Back
Top