Hacker News new | past | comments | ask | show | jobs | submit login
AA A321 takes off after smashing ground sign at JFK (avherald.com)
116 points by howard941 on April 21, 2019 | hide | past | favorite | 42 comments



The avherald reporting is better. And has photos of the damage. Currently they're calling this an accident, not an incident, and per 49 CFR 830.2 that's because of the substantial damage to the wing. http://avherald.com/h?article=4c68c5a8&opt=0

A passenger, who they confirmed was on the flight, says the pilot announced "a major computer failure".

The runway heading and wind conditions mean a crosswind takeoff. The plane would have a natural inclination to weather vane toward the wind (yaw right) and the right wing to have more lift (roll left). Proper crosswind takeoff counteracts this with right aileron and left rudder and slightly higher rotation speed. The left wing should have lifted sooner, not later, the idea being to bank into the crosswind to keep the airplane from drifting; and as the plane lifts off immediately establish a crab into the wind which then permits wings level flight and climb out while compensating for the cross wind in order to maintain a runway heading ground track. Anyway, the exact opposite happened.

I'm not sure which surprises me more, pilot error or computer error. Both seem incredible to me, and this report is eyebrow raising. Dragging a wing on takeoff is extremely dangerous, regardless of the frangibility of airport signage.


"The avherald reporting is better. And has photos of the damage."

The damage pics in the pprune thread are better. You do have to scroll a bit. Here's a direct link to the detailed pics: https://www.pprune.org/rumours-news/620410-aa-a321-takes-off...


It looks like another plane’s jet blast from a 90 degree angle could be a possibility causing it to veer while taking off.

[Speculation by poster Trixie on avherald]


From where? Takeoff runway is 31L, reportedly from intersection KE, therefore 04L isn't a factor, and it puts rotation around intersection PA. No one is going to be at takeoff power over there and not on a runway. Idle power wouldn't do this.

If there's an uncommanded roll before V1, I'd expect aborted takeoff. So it's more plausible this happened after V1, and before Vr, which means the instigator was present for some distance.


From a comment on avherald:

> It is being discussed at American that as the aircraft was at rotation, JFK Ground had cleared a heavy jet into the ramp and he used a great deal of power to get moving. His jetblast, coming from a 90 degree angle to flight 300 on takeoff, is being considered a factor in the roll of the Airbus as it lifted off causing the displacement from the runway and the wingtip drag.

> Powerful taxi thrust could most certainly be an factor. Many years ago I witnessed a DC-10 literally blow a small aircraft up and onto its nose, then onto its back during landing with the thrust of those DC-10 engines powered up while exiting the runway and turning 180 degrees.


Scroll down to these pictures:

https://www.pprune.org/rumours-news/620410-aa-a321-takes-off...

Youch. It took quite a bit of that sign with it.


Found this first on vasaviation (https://www.youtube.com/watch?v=Ca-0Bi2lZhg), which focus on atc communications. It looks like the pilots took 5 mins to report to atc for emergency landing request.


From other reporting it sounds like they didn't know they hit something until left side passenger(s) reported wing damage to flight attendants, who then reported to pilots. Also it wasn't an emergency landing request: no mayday, no pan-pan, their only request was for a particular runway presumably because it was the longer one, and they explicitly denied needing ground assistance.


How can an incident require turning back immediately (and requesting a specific runway!), but not require declaring an emergency? Here's an incident[1] from yesterday where an American Airlines flight declared an emergency after a tire blew. Meanwhile, this aircraft has visible damage to the wing.

1. https://vancouversun.com/news/local-news/flight-from-vancouv...


>How can an incident require turning back immediately (and requesting a specific runway!), but not require declaring an emergency?

Dinking a wingtip is not an emergency. The damage looks alarming, but it's nowhere near anything vital. You want to get on the ground ASAP simply as a precaution, but you've still got full control of the aircraft and the capacity to land normally.

Blowing a tire is definitely an emergency, because your landing could go catastrophically wrong. You need to declare an emergency, because you may need a major response on the ground.


> but not require declaring an emergency?

Having an emergency condition and declaring one on frequency are two entirely different things. By declaring an emergency you're more or less taking control of all surrounding airspace and doing whatever is necessary to achieve a safe resolution of your flight, the tower and ATC both become bound by your declarations as an emergency flight.

If your flight controls aren't damaged, you have enough fuel and/or aren't leaking any, and you're capable of flying the standard arrival pattern it may not be necessary to declare the emergency to gain the privileges above.


Emergencies are pilot's discretion. There's a non-communication exception in FAR 121 operations that doesn't apply here. At least in the U.S. there's no such thing as ATC declaring an emergency for a pilot, and likewise there's no such thing as ATC denying a pilot request once an emergency is declared.

Pilot language in the radio audio suggests the turn around is precautionary, there are too many examples of return to departure airport that are not emergencies. Also, they were offered both 4L and 4R, so it's not like they requested a runway that was not in use.

If there are no control problems, no system warnings, and the pilot sees no damage - on what basis do they declare an emergency? Very different from a blown nose gear tire which is a fire hazard on landing, might grind down the wheel, and some risk of nose gear collapse as well.


> At least in the U.S. there's no such thing as ATC declaring an emergency for a pilot

US ATC can declare an emergency and handle a flight as an emergency in the absence of a declaration from the pilot.

Source - FAA Administration Order JO 7110.65X, section 10. (and conversations with controllers over the years and a flight where I did not declare an emergency, but was handled in every regard as one, including asking souls and fuel on board [an unambiguous indicator])

> on what basis do [the pilots] declare an emergency?

FAR 91.3 gives them incredibly broad discretion and declared emergencies are, despite the name, fairly routine. No pilot would be violated for declaring an emergency under this circumstance.


Just because it looks bad doesn't mean it actually is bad. A blown tire has many more ways of causing a plane to fail during landing then a wing that isn't operating at 100%. As long as all the flight controls are operational and there is no leaking fuel I don't see why it would be considered an emergency.


That’s not surprise.

Aviate. Navigate. Communicate. In that order.


I'm surprised we don't keep video of all take offs and landings. Doesn't seem like it would be overly expensive.


Continuous video surveillance on 3-4 km of runway, from multiple angles, with night vision support, working 24/7? Might not be impossible, but surely not cheap.


I watch landing and takeoff videos frequently on YouTube from only one angle. The lenses are quite large and depth is very compressed but it’s sufficient to understand a lot of things relevant to an accident.

Daytime only would still be a huge improvement.


There's enough plane spotters around Heathrow 24/7 that this is actually probably covered.


It would take a heck of a long lens to have this on video from Heathrow.


Might need a couple of mirrors too...


Many airports do have cctv, ntsb’s channel often include those footage


The NTSB seems to get a lot of work.

https://twitter.com/NTSB_Newsroom


Thanks for that well of information.

This highway crash[0] makes me feel like it's crazy that we still haven't created, and mandated, forward collision avoidance for all tractor-trailer sized vehicles. The people in the SUV never had a chance.

[0] https://www.ntsb.gov/investigations/AccidentReports/Reports/...


Trucks are cut off in traffic extremely frequently. Some trucks do have auto braking and it getting triggered all the time from jerks on the road is a frequent complaint.

That said the NTSB has recommended it be mandatory.


> Trucks are cut off in traffic extremely frequently. Some trucks do have auto braking and it getting triggered all the time from jerks on the road is a frequent complaint.

Yeah, false positives would be a giant source of friction. Do you know what types of sensors the existing systems use?


Nope no specific knowledge, just a decent amount of trucker griping. The stopping distance of trucks is much longer so the triggering distance on these devices must be correspondingly longer as well. A lot of cars don't even realize they cut off the truck because they can stop so much faster.

A trucker can assume that a car cutting them off with open road in the front won't slam their brakes. These devices can't.


I'm confused as to why there would be a sign anywhere near the flight path of an airplane.


Not the normal flight path of an airplane, anyway.

I don't know which kind they hit. But based on what's on the left side of runway 31L, I suspect it was a runway remaining distance sign, those are all black, and around 5' tall, and vaguely resembles the passenger photo of something black embedded in the wing.

Edit: Comment #77 in this forum thread is interesting if a true account. https://www.airliners.net/forum/viewtopic.php?f=3&t=1420137&...


It's a runway marker sign. They are pretty low, and in the grass, and have breakaway mounts. This sort of thing: https://blog.klm.com/assets/uploads/2015/08/Signs_7.jpg

The aircraft would have been way off center of the runway, with the left wing literally dragging the ground.

JFK runway 31L is 200 feet wide, an A321 has a wingspan of 112 feet. The wings are 12 feet from the ground at the fuselage, and maybe two feet higher than that at the wingtips. The photos[1] show heavy abrasion under the left wingtip, presumably it was dragging the ground.

[1] https://imgur.com/a/rKu8cUC


Also if there’s wind and the gas in the shocks is a bit low you can easily hit them.


Fron TFA, it was taxiing when it hit the sign.

Edit: Disregard this, AV Herald links provide more info.


This appears to be incorrect. Reports state damage was sustained during takeoff.


This is the AVherald article about it, with more details: http://avherald.com/h?article=4c68c5a8&opt=0



Expensive accident. Glad everyone is safe!


Does this mean the plane "rolled" slightly to the left while still on runway, and hence the left wing was able to hit that sign? Did the pilot use the ailerons while on runway?


Unverified video taken on the ground after landing. https://www.airliners.net/forum/viewtopic.php?f=3&t=1420137#...

It looks similar to other photos, and if authentic yes it suggests a significant roll to the left. From the available information I can't assess whether the plane was on the ground or airborne, either are possible.

Aileron is normally used for crosswind takeoffs, to lift the leeward wing sooner in order to "lean" the plane into the crosswind to keep the plane from drifting from runway centerline. Essentially the opposite happened in this case. The airplane must have both drifted to the left of the runway center line (not out of the ordinary, in particular if it's gusty) as well as left wing dipped.

It's possible this was a suddenly gusty takeoff. I'm not sure to what degree A320 series can do auto takeoff or computer assisted takeoffs, and what crosswind limits they presumably have where a pilot has to do the takeoff manually instead. Maybe the pilot's suggestion of a computer problem is related to that, and they subsequently took manual control.


Aileron is normally used for crosswind takeoff

Airbus crosswind takeoff recommendation is no lateral stick input. The sidestick is a rate controller, so a lateral deflection commands a roll, not a specific control surface position like on non-fly-by-wire aircraft. And you run the risk of deploying spoilers.


The original article article URL before the change to avherald contains a post that addresses this: https://www.pprune.org/showthread.php?p=10450572

Obviously we'll have to wait and see what the actual story is, whether this is really old bad tribal knowledge, and if the operating manual contradicts that, and whether pilots are getting the proper training and consistently applying it in practice.

I'm an opinionated and biased pilot and former flight instructor, but not an aerodynamic engineer. Spoiler extension does not increase the effective angle of attack of the wing, whereas a dropping aileron does. And therefore off hand, and without benefit of an Airbus type rating, I don't really care if the actual control surface is a bit of right spoiler instead of left (down) aileron, all I care about is that the windward wing does NOT lift off first because that will foul up the ground track on takeoff, and that's simply not proper technique.


I'm not typed either, for the record. Current flight instructor and my boss is a check airman at a major--that's about all I've got to offer. :) As I understand it there's a potentially dangerous pitch-up moment that occurs if the roll spoilers deploy on takeoff. (Don't ask me why!)


I am not sure about this. Control has several laws and for landing it is afaik not a rate controller but direct input. It could be the same for take off.




Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: