With a little free time, we’re catching up on the Asiana accident.’
Ernie Arvai sent us this link from The Atlantic which references several postings on a professional pilot forum, PPrune.
Several readers linked or sent us this article which is thorough enough that we’re elevating it to a prime position in a posting.
This article details some of the debrise found in the water and some more key facts about the final 1.20 minutes of the flight.
This article outlines some of the spinal and neck injuries incurred. These may be studied for future seat designs to increase safety.
The outcome-only two dead (as yet) and several still in critical condition is being hailed as testimony to aircraft safety and luck. There is no question aircraft accidents are more survivable today because of the safety advances. But if the 777 had come down on the seawall any further forward than it did, the outcome could have been far worse. This is the luck part. Professional flight attendants and passenger heroics also played a role in survivability. Itäs tragic that one of the two deaths appears to have come from a rescue vehicle accident.
I think I will wait for the official accident report.
If you want hard facts, with intelligent comments, see http://avherald.com/h?article=464ef64f&opt=0
Thanks iPuck. There are some pictures I have not seen before.
It appears that the lead flight attendant was very much a hero in this, she seems to have known and done more than the pilots did in the post crash evacuation. In the interviews she did in the Korean media, and reported here
She did a great job. The Pilots??? Well, they don’t appear to have done much. The co-pilots did help significantly, but one must ask where were the pilots? In her narrative they didn’t appear much. As the Lead FA, she was the last off the plane? Again, where were the pilots? It’s supposed to be their job to be the last one off. The more that comes out about the pilot’s performance, the more questions arise.
The note about the possibility that the rescue vehicle may have run over one of the two girls is so sad.
Howard, excellent story. But I will reserve judgment on what the pilots did, or didn’t do to help the passengers after the crash. At this point I doubt we know anything of the actions after the crash.
quality of the article “elevated to prime position” is excellent and a must read for everyone
As an A320 pilot for a US major 121 carrier I will leave those who know the 777 to speak to the technical issues like the FLCH trap and I think The Atlantic article does an excellent job of summarizing how this crew could potentially of found themselves getting so slow.
What I want to highlight is the fact that SFO and may other airports in the US have a long and documented history of “slam dunk” approaches where pilots are often kept well above the normal glide path and then asked to slow and descend rapidly. These contradictory requirements (in general you can go down or slow down but not both) force us into automation modes and to use measures that are outside the norm and lead to unstabilized approaches.
The industry of course generally requires approaches to be stabilized by 1000′ AFE but data shows that crews routinely push the envelope in an effort to make the landing work and avoid a go around. That aspect is one that is getting a lot of industry attention right now.
What isn’t getting any attention is the fact that ATC routinely creates through the way they handle traffic these unstabilized approaches and that pilots are being pushed into doing things they recognize as being contrary to best practice in order to make the approaches work. SFO is a well-known problem area and another good example is the 18R approach into Orlando International (KMCO). On that approach ATC routinely at the last minute and without warning (unless of course you fly there regularly and know it’s coming) issues instructions to cross the ORL VOR (co-located with Orlando Executive Airport) at 2500’ instead of the 2200’ the final approach fix on the ILS calls for when issuing the approach clearance. This 300’ may not seem like much but especially if a crew is unprepared for it very easily results in an unstabilized approach.
Pilots have complained for some time about ATC’s poor practices and in my opinion Asiana is the inevitable accident that was sure to result from this trend of forcing pilots into unstabilized approaches.
I will add an interesting anecdote. After on particularly bad vector into KSFO with multiple slow down and go down requests I got a phone number for the TRACON and called to complain. The supervisor apologized and promised to review the tapes of the incident. During our conversation he mentioned that they had been forced to implement a rule requiring Asian carriers to be vectored onto final at or outside a proscribed distance and always below the glide slope. It would seem from the preliminary data that rule was violated as well.
How can a 777 become so slow so close to the ground without bells ringing?
The stick shaker began at 4-5 seconds before the first impact. A go around was attempted after that.
hypothetically if that was the warning bell keesje is referring to, how could it be of any use if it sounds that short before impact? In other terms, what is your point?
The airplane was providing a warning of an impending stall. The airplane did not know it was seconds from crashing.
How is it possible for TWO trained pilots, to sit there and NOT realize that they
were below the glide-slope AND below minimum speed, before it was too late!
Very bad piloting, that’s about all that can be said!
From my own design and operating experience with large ground-based tracking radars, (particularly if the radar is situated close to a sea/land interface) there may be a problem where cold air can collide with warmer air.
Between the cold Pacific Ocean’s Humboldt current that flows north near SFO and the warmer water of the Bay, on odd occasions there can exist a “ducting” phenomenon of electromagnetic waves (certainly for C-band radar frequencies with which I am familiar and perhaps also for visible frequencies) that can distort any transiting electromagnetic-wave propagation to some extent.
Could such a coincidence have confused the pilots’ visual perceptions even in some small extent?
There is amateur video (shot at a distance) online that captured the last seconds of the plane landing/crashing, including what appeared like the plane flipping over (actually not) a couple of seconds before it came to its final resting position.
From the sound on the video, it was a whole bunch of people who witnessed the whole thing, albeit at a distance.
The amateur video is included in the cnn.com story here:
Sorry you have to watch the commercial when the video starts.
Quoting from the story above:
“When the aircraft hit, it spun 360 degrees. An oil tank ruptured and leaked fuel onto the plane’s right engine, starting a fire.”
This is in agreement with one of the comments made on another blog piece posted here a couple of days ago.
Hundreds of interesting (777) pilot posts on the topic.
Things that came up
– Korean seniority culture could have played a role (pilot was “senior” to captain)
– Pilot was a very experienced a320 (“the speed control and protection is very reliably looked after for you in the 320, so the speed scan becomes low priority”)
– Human error is seen as the most likely behind this near disaster (what were the 2 crew members doing?).
The auto throttle operation now has full attention, apparently the pilots thought is was on.
In my view, there are serious inconsistencies.
I have never flown an aircraft where the stall warning did NOT sound prior to stall.
The question therefore is how the stall warning happened so late, just a few seconds prior to impact, and yet according to NTSB the impact speed was just 103kts?
The approach speed should have been 137kts. Anyone shed light on the stall speed of the 777 in landing configuration?
My feeling too (without diving into it).
The aircraft “knew” the aircrafts height, angle of attack, minimum approach speed, vertical & actual speed. Why wasn’t there a horn at e.g. 10% below minimum approach speed?
I can some interesting discussion coming up. Should the aircraft warn a pilot when he is acting dangerous or is the crew in full in control & not to be disturbed. Should the aircraft be allowed to prevent a crash just before it become inevitable?
All we know at this point in time is about the stick shaker. We don’t know about if any flags, annunciator warnings, lights, horns, etc., were issuing warnings, and not noticed, or ignored by the flight pilot and instructor pilot. Of course any or all of these can be turned off or disabled by the crew.
I know of no airplane that can, on itself, prevent a crash. There are systems used by Boeing and Airbus that reduce the possibility of a crash, but nothing that will completely avoid a crash.
The B-777 is fully FBW with flight envelope protection (FEP), just like Airbus. But Boeing is unlike Airbus in their FBW systems by allowing the pilot to override the system.
So this goes back to the pilots and not the airplane design, as you are inferring, keesje.
I thought it to be a rule of thumb that stall speed in landing configuration times 1.3 is the approach speed. Can anybody confirm or dismiss this?
Bob, under normal conditions, approach speed is 1.3 times stall speed. For the 777 normal landings are at flaps 30 and, of course, gear down. 1.3 VS gear and flaps down is the landing reference speed VREF.
VREF is always displayed on the airspeed tape during approach, so either pilot can tell at a glance if the airplane is too fast or too slow, Any pilot that ignores the speed tape long enough for the airplane to be 10 kt too slow on final approach should be fired immediately. To allow deceleration to minus 20 or 30 kt – don’t even think about it.
According to some reports VREF was around 130 kt. If the stick shaker activated at around 100 kt, that would have been normal.
The instructor pilot said in his interview that the plane was flying slightly high when it passed 4,000 feet and was headed toward the runway, while at 500 feet he began realizing that they were flying too low. He saw the precision approach path indicator(PAPIs) lights, or a set of four lights beside the runway, were not two white and two red as they ideally are when landing–instead the set was three red and one white, which indicates the plane is flying too low.
Do we know if the full on board complement of 4 pilots were in the cockpit ?
With the presence of additional persons probably being of ambivalent value.
The TK crash at Amsterdam and QF32 come to mind.
Uwe, I read somewhere recently there were 3 pilots in the cockpit, and the 4th was in the cabin. I do not know if that is reliable.
For a fascinating commentary on pilot training at Asiana and Korean Air, go to
Read both comments by “boomer depp”. Try to ignore anything in the thread that’s snarky nasty or incoherent.
Read ’em and weep, not just for the families of the two fatalities, but also for the passengers who have had their lives ruined by this accident
NTSB Chairwoman Deborah Hersman said that the pilot and instructor seated next to him stated they had programmed the auto-throttles, which operate like cruise control in cars, to maintain the target landing speed of 137 knots. At 200 feet altitude, the pilots realized the auto-throttles weren’t maintaining speed.
Investigators found the devices set in an “armed” position, meaning the throttles were ready to be engaged. But it’s unclear if the pilots set them properly so that they would maintain the proper air speed.
“We need to work to understand what the different modes are and what the crew understood,” Hersman said. She said investigators are working with Boeing, “which is providing us with information about systems and how they are designed.”
She cautioned against jumping to conclusions. But, she added: “Let me be clear: The crew is required to maintain a safe aircraft and that means they need to monitor. We have the flying pilot and two other pilots, and they need to monitor functions. One of the very critical things that needs to be monitored on landing is speed.”
As noted above VREF is always displayed on the airspeed tape during approach, so either pilot can tell at a glance if the airplane is too fast or too slow. It doesn’t matter if Otto Throttle was armed, on, off, or sitting in the cabin reading a book. Ms Hersman was absolutely correct: “[the oilots] need to monitor functions. One of the very critical things that needs to be monitored on landing is speed.”
How about Otto Pilot?
ROTFLMAO……..for those to young to understand Scott’s post, Otto Pilot was the inflatable B-707 auto-pilot in the 1980 comedy movie “Airplane”.
“Airplane” is one of the funniest movies of all time.
KC, I have to agree. You also gotta love Barbara Billingsley as Jive Lady; “Chump done wan da heyp, chump done git da heyp”
The skewed reference to “Airplane” was deliberate. Vector Victor?
And stop calling me Shirley.
Get the clearance, Clarence, roger, Roger……
The scene where Julie Hagerty (as Stewardess Elaine Dickinson) has to manually reinflate Otto Pilot after he springs a leak is just priceless. After he is reinflated, he smokes a cigarette.
I wouldn’t be amazed if an audible warning existed in the aircraft if it was e.g. below 10% the local stall speed. Doesn’t seem like rocket science to me. Why not? The extensive combination of training, procedures,aircraft and airport systems failed. Thats for sure. (slightly provocative, I know).
10% below the local stall speed, a warning is no longer required.. I meant approach speed..