http://www.foxnews.com/us/2013/07/06/boeing-777-crashes-at-san-francisco-interna...

It was human error not the airplane that is good

ILS was inop and there was a NOTAM released. There is also some speculation that the VASI lights were inop as well. Weather clear with 10+ vis/

http://www.youtube.com/watch?v=XLYeUbeyfOg

We armchair pilots do not have the right to speculate NOR should we. I've seen most all the footage and all I can say is the B777 is the MOST sturdy aircraft I've seen to date. To come to rest after that with so few casualties is nothing less than amazing. I also refer to Heathrow!. AND as with that incident we still cannot eliminate fuel (iceing) contamination or what ever.

I have to agree with Bruce of Lost in Isaan:

The 777 is truly a sturdy built aircraft. The amount of structure that remained at zero motion was remarkable.


The following is speculation

Here is a printout from a site that records flights from radar information.
http://flightaware.com/live/flight/AAR214/history/20130706/0730Z/RKSI/KSFO/track...




Interesting data here.  Less than 2 min to landing, he was at 169 kts/194 mph passing 1400', vertical speed 1380 fpm down.  Descent rate remains at 1300 fpm or more until 600' above the ground with decreasing airspeed.  Airspeed continues to decay to below normal as descent rate decreases to a more normal value.  Looks like the nose is coming up but without power increase.  Then at 100' above the ground (water) airspeed is down to 109kts, looks like a big pitch up at low airspeed.  Likely they were in a stall condition prior to touchdown.  Fits a classic high unstabilized visual descent with an pitch overcorrection near the ground with little to no power increase leading to a stall.  Just my opinion from the raw data.

---

I flew a few approaches in the CS 777 to see what it would look like from the cockpit and the ground. I set a landing weight around 400,000# which gave me a Vref of 129 kts. After flying several approaches and doing screen captures of the approach I came up with these two shots.

Here is the plane in landing configuration. I have less than the necessary thrust set to make the speed bleed off to near stall.
The speed in this picture is 108 KTS and a vertical speed of -420 FPM @ 100 feet above the runway, with 10 degrees nose up.
I continued to raise the nose to touchdown where I obtained a pitch angle of almost 15 degrees and in stall buffet the last 50 feet above the TDZE.



Here is a snap of the plane just prior to touchdown.



In this case, I was able to land on the runway in stall buffet and just scrape the aft part of the plane. If I had been just a little lower on this approach the gear and the aft of the plane would have hit short of the runway. It looks like, from the cell phone video and passenger reports, that the plane actually skimmed the water prior to hitting the rock wall.

Now, I have never flown the 777, but the other Boeing planes like the 747 and 767 all have pretty much the same auto-thrust system. The throttles are driven by a slip clutch when A/T is engaged. You can override the throttle movement while the A/T system is still engaged. This is something I would do many times to make small corrections to the speed. Since the plane was high and fast, I wonder if the pilot disengaged the A/T during the descent in order to get the plane down into the slot and then "forgot" to re-engage, or thought they were engaged. He could have just kept his hand on the throttles and held them at idle, but if he moved his hand off the throttles to maybe do something else the throttles would have increased if the speed was below what was dialed in on the glare shield speed window. In any event, I can only guess that the A/T were not on during the last portion of the flight and that resulted in a speed bleed as the nose was slowly raised during the last part of the approach.

BTW, during several of my simulated approaches, I kept the throttles back during the last portion of the approach in order to force the stall around 100 feet above the ground. Each time I got the speed down to around 106 KTS, buffet would occur. If I applied full thrust at that point almost every time the plane would make a successful go-around.

Lou

Uploaded with ImageShack.us

Thank you for the analysis Lou. Was going to go through the same exercise myself this morning, before doing so I spent quite some time viewing Fred Hayes's video. Looks like they were dragging the tail through (or just over) the water for some 18 seconds or about 10 plane lengths, between 550 and 650 meters, prior to hitting the sea wall.

Going to have an interesting few hours replicating that. Crash detection OFF!

To set the record straight, this is how training and qualifying go in the airlines today.

If you are already a qualified line captain, or even a new pilot to the operating certificate the training will generally follow this way.

First, you will go through several weeks of ground school and pass a written test.
Second, you will have an oral exam in all the systems and procedures pertaining to the aircraft and the operation.
Third, you will get a specified numbers in both a systems trainer called a CPT or cockpit procedures trainer.
Fourth, you will go through a number of hours in a six axis motion flight simulator that will include all parts of the normal, abnormal and emergency flight procedures that will end with a full check ride and a LOFT (line oriented flight training). If you pass all of this you receive your rating in the plane. The only thing left to do is get a specific number of hours of flying with a line flight instructor/check airman which on international types of flying would be at least two round trips and end with a line check ride.

These are the minimum things that need to be accomplished, if a pilot needed more time, it could be given.

It has no bearing as to whether this was his first landing in the real plane or if this was his first time in KSFO. Only very different airports such as high altitude or restricted terrain like PARO or La Paz would you get a check pilot to go along on your first time flying there. KSFO is a very ordinary large airport.

Even with the ILS G/S inop, the pilots could have built the approach in the computer and flown it with computer guidance.

I just have to wonder what the check pilot was doing through out the whole approach. Sure, you have to let the "student" make some mistakes to see if they will correct them, but once you get into the low regime that freedom will have to stop and the check pilot takeover. As a pilot for TWA and American, the rule was that the plane MUST be configured for landing before 1,000 feet on the approach. At 500 feet the non flying pilot (NFP) would call out speed & rate of sink. For example: The NFP would announce - 500' speed 137, sink 700. If the sink rate exceeded 1,000 FPM a go-around was required. The computer at 500 feet will call "sink rate" at around 1,500 FPM.

In this case the sink rate was not a problem, but the speed bleed was legion. To have the reference bug set at 137 KTS and see the airspeed needle at 106 KTS is beyond my understanding. What was the check pilot doing?

I think the Chairwomen of the NTSB, Deborah Hersman is very smart and in due time we will see just what happened.

At TWA our motto was "Fly Aggressively Safe."