Unless the guy had been flying small prop planes until now to get 10,000 hours - I seriously doubt there are too many large passenger jets where approaching well below 137knots is acceptable.
I’m not a pilot, so please correct me if I’m off base. From what I read and see on tape it seems as though the plane was coming in too low and slow so they applied more power. The stall warning came a few seconds later. Is it possible that the pilot pulled up the nose too much when he saw he was coning in short? That would have decreased speed even more, put the plane into a stall, and made it land even shorter than it would have anyway. The nose up position would also have caused the tail to hit the ground so heavily. So rather than landing short but in a fairly conventional angle they landed even shorter and at a greater angle.
Also, did they apply full thrust? Would that have caused the plane to lift up so much after it hit? Or was that just a “bounce”?
This (subordinate flight crew members being reluctant to present any kind of challenge to the captain’s judgement) was a problem for Korean Airlines until the late 1990’s, when they specifically began training flight crews in the tenets of crew resource management and worked hard to teach subordinate crew members to voice their concerns in explicit, unambiguous (but courteous) terms, and senior crew members to accept and consider that sort of input without feeling that their authority or status is being challenged.
Actually as I understand the history of CRM, this used to be a problem for all flight crews; it’s just that KAL was a late adopter, and Korean culture also happens to be exceptionally deferential toward authority figures.
Around the same time, KAL began insisting that its cockpit crews be fluent in English, which helped cut down on miscommunications with ATC. These measures helped cut their accident rate from one of the worst to one of the best.
I wonder whether Asiana - Korea’s other airline - has taken the same measures with their flight crews.
From the tapes I heard of communication with ATC their English didn’t seem that good.
I will bet one wooden nickel that it comes down to a seniority issue in the cockpit like that of flight 007. Given that this was another Korean aircrew, I’m going to bet that someone was too intimidated to interfere or ask the right questions because of a seniority issue.
Thanks.
I found this in a report (emphasis mine), in case anyone else missed it as I had:
ISTM that the copilot should have caught the captain’s mistake(s) and mentioned it (them) to him. I was unaware of the defrence-to-seniority issue amongst Korean aircrews until I read about it in another thread.
Man that is crazy!
I’ve been reading the thread on Pprune and the issue of deference to authority has been discussed there at length, but unfortunately seems to have turned to beating on Asian culture.
One aspect I haven’t seen mentioned is the wind: was there a sudden shift in the wind? A 10 kt headwind turning to a 10 kt tailwind means a drop in airspeed of 20 kts.
I saw it reported on airliners.net that this was the first time the other pilot was flying as a certified trainer.
So, you have pilot A who’s inexperienced in the plane, and pilot B who’s inexperienced providing training. Pilot A has more experience overall but hardly any in the 777, while pilot B has substantial experience in the 777 but less overall, and no real experience providing training.
It might get more interesting than this, but it seems like a familiar story. Pilot A screwed up because he was new to the plane and pilot B, not wanting to offend the “more-experienced” pilot A, waited too long to take corrective action.
I’ve worked on avionics development for a few wide-bodies, although not the 777. It’s my understanding that everything made since the 90’s has a TOGA* button. I’ll defer to Richard Pearse of course, but wouldn’t the acting pilot simply have pushed TOGA when they decided to go around? Is it possible that CFIT** was unavoidable at this point, even with the airplane conducting a max go-around? I wonder if the pilot(s) grabbed the wheel in a hopeless attempt to prevent contact with the ground? (Thus causing the extreme nose-up angle) Just my WAG, of course.
*TOGA = Take Off Go Around, ie. instructing the plane to maximize climb starting right now. At least one aircraft (MD-11) has inward facing actuators between the throttle levers to detect a stressed pilot is “squeezing” the levers, and will initiate a go-around if it happens. I can’t confirm this design was for reading a stress reaction, but it sure seemed that way every time I flew the sim, and that was the reason given me by a senior designer. I’ll also freely admit I’m not a flight controls expert, so maybe TOGA as I understood it wasn’t for emergency go-arounds. Hopefully the experts can set me straight.
**CFIT - Controlled Flight Into Terrain. Normally this would refer to an inattentive crew simply driving the plane into the ground, but I believe it would apply to a case where physics prevented the plane from regaining a positive climb rate before impacting the ground.
The TOGA buttons do different things depending on what automatics are on. I can’t speak specifically to the 777 but typically if the autopilot is off and the auto-throttle is on, the TOGA buttons will command max thrust but you still have to hand fly the procedure. If the autopilot is on then the autopilot fly the procedure for you and the auto-throttle will give max thrust. If the auto-throttle and autopilot are both off then, in my aircraft at least, the TOGA buttons don’t really do anything except tell the engine computers to trim the thrust to the required value, you still have to move the thrust levers forward into the trim range for the engine control system. So the TOGA buttons aren’t a simple get out of jail free card, you have to know what you’ve got to know what you’re going to get.
I’m pretty sure that 1.5 seconds prior to impact is far to late to be commencing the go around though, regardless of what technique you used.
Was CRM an issue with KAL 007? My understanding that the issue with that flight was that a Soviet air force jet shot a missile up their tailpipe. Was there some CRM issue that resulted in 007 flying over Soviet airspace?
He’s talking about a different flight I guess. There are a few to choose from.
KAL 801, a 747 crash in Guam, is a much better illustration of the issue.
It’s starting to look like pilot pooch-screwing but a full investigation is important to rule out any other causes or factors. We don’t want to zero in on the obvious only to miss other important facts.
^ What Mr. Pearse says. As noted, at some point every pilot (captain or co) had to make his/her first lading in a particular type of airplane. Standard practice these days is lots of simulator time prior to ever taking the controls of a real airplane. At some point, though, that first/second/nth landing has to happen. While that may be alarming to the flying public who aren’t pilots, it’s standard, normal practice and as is often repeated flying is the safest form of transportation we have these days. There’s a lot of prep that goes into training a pilot before he ever sits down in a new cockpit and the safety record indicates that it works. We can hope to continue to improve training and safety, of course, but I don’t think the current system is a problem. Individual pilots might be a problem, but not the overall training system when properly done.
Eh, there’s not any easy way to work that out. If you’re talking about short-hop flights (which this wasn’t) then 47 hours might be 45 landings. If you’re talking about long-haul flights (which this was) it might be as low as 4. Heck, I think there are some routes where 47 hours might have less than that.
Try to break away from the old captain/co-pilot distinction. Accident reports more and more are moving to the pilot flying/pilot not flying distinction.
Now, let me try to explain this, hopefully without screwing it up. “Captain” is a title based on seniority within a company, how long that pilot has been employed by that airline. So if a captain from one airline moves to another (maybe due to employer bankruptcy, whatever) he starts over on the bottom of the seniority ladder. So it is entirely possible that a co-pilot could have 10,000 hours more than the official captain of the flight, and it has and does happen that way.
There is also this notion out in the general public that the captain does all the flying, or is somehow inherently more experienced/skilled/competent. This is not the case. Common practice is for one of the pilots to do one landing, then the next landing the other is flying. This keeps everyone’s skills fresh and (one hopes) improving. Also, any pilot in the cockpit of an airliner has demonstrated being a competent pilot for quite some time and thanks to simulators and modern should be entirely qualified to fly that airplane.
In this case, the captain of the flight was captain because of seniority within the company. In fact, three of the four pilots aboard were captains. The co-pilot wasn’t a captain but he did, in fact, have more experience in that type of aircraft than the flying pilot at the time of the crash. That’s probably why the two were paired up. However, as mentioned by others, the PNF (pilot not flying) was inexperienced as a training pilot. All trainers, though, have to have a first flight where they are training. Usually this is not a problem as it is exceedingly rare for any commercial pilot to have no teaching experience whatsoever (most have some flight instruction experience in their career) and this guy probably had some, but likely not in the context in which he found himself.
Pretty good explanation, particularly for a non-pilot. My understanding (which is not perfect and subject to change with more information) is that the stall warning came before the corrective action. I think I’ve been hearing 7 seconds, but they didn’t apply power until 5.5 seconds later, which seems like a significant delay to me. Yes, it’s possible the pilot pulled up the nose too far, but that might also have been caused by increased power causing the nose to pitch up, or the tail hitting the breakwater causing it, or a combination of factors.
Yes, they did eventually apply full thrust but too late to do much if any good. The lifting of the airplane was, more or less, a bounce.
Yes, that could be an important context for this accident. Such pairings do occur rather regularly and usually work out but it may be that they reduce the safety margin.
While ideally all pilots flying would be super-experienced at some point you have to train new people. At some point, even the most experienced pilot had to have that first few flights/landings.
In very short, the captain is believed to have mis-set the flight waypoints, which caused a slow, cumulative drift off course until they overflew Kamchatka Island instead of staying 12-15 miles offshore. The Soviets panicked and (by their own admission, much later) over-reacted and shot the plane down.
It’s believed that the flight crew was too intimidated by the Korean culture of respect to second-guess the captain’s settings, even when navigational errors began to multiply.
Question for the pilots and/or aviation admin types: When computing the number of hours a pilot has in an aircraft, does one only computer the time the pilot is in control or does one enter the total time the pilot is in the cockpit?
The reports I have seen have it the other way around. At 8 seconds there was a voice in the cockpit calling for more power and the black box shows that the throttle was advance. The stall warning came afterwards.
Excellent post BTW, one of the fun things about the SDMB is that there is always someone who actually knows something about the subject being discussed.
That’s more complicated a subject than it should be. For a plane that requires only one pilot, then only Pilot In Command time counts - but you can accumulate that if you’re acting as an instructor or safety pilot for someone else who’s at the controls. For a plane requiring two pilots, like the 777, then you can log Second In Command time too, but in a different column.
I think the bigger issue is that not all hours are alike. An hour cruising over the ocean, half asleep, counts as much as one fighting an approach into an unfamiliar airport with bad weather, etc. Hours flying different aircraft types under different conditions provide more useful experience than the same hour over and over.