Monday, January 17, 2011

Jackson Hole Overrun

I referred yesterday to the American Airlines runway excursion at Jackson Hole, Wyoming. That's looks like a tough, could-happen-to-anyone situation. The high airport elevation makes your gut instincts wrong, because as a result of low density air the airplane is moving faster over the ground than the airspeed indicator shows. The visibility means that you're on the gauges until decision height, then, drifting in a crosswind, you look up, find the runways lights off to one side, and land on the pavement that you trust is outlined by the lights, with everything reflecting off the swirling snow. The snow doesn't even look like snow. It looks like some weird flashing thing, just stabbing at you. Honestly, passengers who are perfectly familiar with the terrestrial sort of snow ask "what is that?" But you can't sit there and stare mesmerized like the Enterprise bridge crew. You have to land, make sure the thrust reversers have deployed, brake, trust the antilock, keep straight with no visual references and uncertain braking conditions and by that point all the formulae and calculations and runway friction data are academic. You are doing your job and just hoping that the rate of deceleration you're getting is such that by the time vee equals zero, half eh tee squared is less than runway length minus touchdown point. The Boeing 757 is slowed after landing by thrust reversers which direct the engine thrust forward, spoilers, which both provide aerodynamic drag and reduce the lifting power of the wings, increasing the weight on the wheels so that the wheel brakes are more effective.

The chart I have for the Jackson Hole runway 19 ILS approach shows a standard 200' decision height and what Canadians call an advisory visibility of three quarters of a mile. That is, one can reasonably expect to be able to see to land on the runway with that visibility and a ceiling above 200'. Reported conditions at the airport were a ceiling of 400 feet, light snow and visibility of 1 mile. Winds were 10 knots from 240 degrees. So the pilots expected to be able to land, and they did, but what went wrong?

I don't know. It's a sexy looking airplane, but I've never flown anything remotely like it. The information I have tells me:

The FDR data indicate that the aircraft touched down at approximately 132 knots. At touchdown, the air/ground parameter changes to "ground" for approximately one second and then switches to "air" for approximately ½ second before changing back to "ground" for the remainder of the recording. The thrust reverser discrete parameters on the FDR indicate that approximately 18 seconds elapsed from the time the reversers began moving until they were fully deployed.

I don't know the recommended touchdown speed for the aircraft at its actual landing weight, but 132 knots does not seem out of line. Without more knowledge of this aircraft systems I naïvely assume the air/ground parameter information means that the aircraft bounced slightly. This appears to have reset the thrust reversers and spoilers the same way a go-around would. Both are armed on approach and then activate automatically when the airplane sensors show that the main wheels are on the ground. This discussion suggests that wheel braking is restricted to half until sensors report that there is weight on the nosewheel. More informed discussion elsewhere doesn't mention the possibility of a bounce, and it's hard to see in the video. It does show very slow deployment of reverse thrust. Perhaps it was a landing gear sensor that "bounced" rather than the physical airplane. There also seems to have been a maintenance problem with the autospeedbrake (i.e. spoiler) handle in the cockpit, something that I'm sure will be scrutinized closely.

Look in the comments for possible contributions from readers more knowledgeable about this aircraft.


amulbunny's random thoughts said...

Interesting video and assumptions.....

Have you had a chance to see Discovery Channels program Wild Alaska about ERA Aviation. I find it more interesting now that I've been reading your blog and I remember bits and pieces of your winter tales.

My cruiser just flew a 757 from FLL to ATL, and will hop in a 767 from ATL to LAX later tonight. Home after 15 days on a Central American study cruise. Lots of tales to tell there.

Anonymous said...

130+ touchdown speed is a tad high, but given the altitude it's probably not excessive.

What surprises me is the use of a B757 for a flight to Jackson Hole. It may be normal, but I was under the impression it wasn't certified for anything larger than a B737.

B757 spoilers are inboard, not visible from the window the video was shot through (as hinted in the video as an option).

It appears the aircraft touched down short, maybe anticipating a long brake distance on a potentially slippery runway.
If it indeed bounced, causing the automated systems to reset, TR, spoilers, and wheelbrakes would have to be manually engaged again, costing a few seconds.

So what we seem to have is a combination of a somewhat rough landing on a fouled runway that's already on the short side for the aircraft type and weight causing late braking, with an overrun as a result.

I'm almost tempted to install Microsoft FSX again and give it a try, but I'm terribly out of practice there (another good reason to install it again, maybe).

Anonymous said...

I was in Jackson Hole exactly 2 years ago and the terminal was full of commercial aircraft larger than 737's ... it was neat because there was no visible pavement on either the tarmac or runways - they were entirely snow-covered... The most surprising thing about the postscript to this event is that the airline apparently removed the black box and downloaded the data prior to handing it over to the NTSB - I'd have thought that was highly improper.

Sarah said...

130+ touchdown speed is a tad high, but given the altitude it's probably not excessive.

Well, no. Indicated airspeed is affected by the same air density that affects the wing. Increasing airspeed with high density altitude is not a good idea.

Sorry for the comment pedantry.

I have no comment on 757 performance @JAC, never having been in the pointy end - and won't guess because a lot of the facts are not in yet.

Aviatrix said...

Sarah is correct: the 132 kt touchdown speed cited is an indicated airspeed, so for the same weight should be the same at any altitude. It LOOKS way faster at altitude though, because you have to go faster over the ground to pass the thin air at the same rate.

Anonymous said...

misread it then, was under the impression they were listing groundspeed on touchdown, not IAS.

Aviatrix said...

I have some more comments on this from an airline pilot, not least of which is that fact that thrust reversers are not armed before touchdown but must be manually deployed. I'll put it in another post.

5400AirportRdSouth said...

The specualation I was reading pointed to the fact that the reversers do not actually fully deploy at a point when they should have ( manually or otherwise ). In fact, if you watch them in the video, you can see them open up slightly, less than an inch, sit there for a few seconds and then close and finally fully deploy.

I'm going to speculate that their was a mechanical issue with their deployment and by the time it was realized by the flight crew, and recycled, it was too late into the landing roll.

It gets more and more interesting every day in the amount of "oh-firetruck " moments that are captured routinely by cellphone video cameras.

Michael5000 said...

I'm just kind of dazzled by the intelligence of the analysis in a TV-like format. Who knew that words in combination with video could be... smart?