British Airways flight 492 aborted a landing recently. Someone on the ground recorded this video of the climbout after abort. Note the massive roll oscillations, which appear to be very repeatable in terms of amplitude and frequency, and are not accompanied by any noticeable yaw/pitch deviations. It sure looks like the pilot is just sawing the yoke back and forth.
And then there’s this video from a passenger on the same flight. You can’t see the aileron out at the wingtip, but the raised spoiler tells you when the pilot is applying large left yoke input. Left roll movements begin with the pilot’s left-roll input, and end when the pilot ends his left-roll input.
What the hell is going on here? Is this just the world’s worst pilot, not understanding that he needs to start backing off of the huge roll input well before the plane gets level? or is he deliberately screwing with the passengers just for his amusement?
ISTM you’d want to start applying large roll corrections long before such large bank angles are reached. We can’t see the right wing to know for sure, but I’d bet large sums of money that the pilot is applying right roll input until the moment max right roll is achieved (at which point we see left roll input being applied).
My point is that the violent oscillatory roll behavior seems entirely driven by the pilot’s control inputs, rather than any external turbulence. Wondering if pro pilots agree with that assessment, and if so, why this pilot might have been flying like that.
I’m not sure we can conclude from that short sequence that the whole thing is caused by control inputs. But I think he’s overcontrolling. We see a roll right (wing goes up), and he seems slow to react to make a left-stick control input to correct it; since he’s late I think he then falls into the trap of making too large a corrective input, and then he’s too slow to remove the input as the aircraft levels, causing a roll the other way. That would be classic over-controlling. But that’s really a beginner thing, something you’d not expect from a pilot with even minimal experience.
But I’d like to hear from one of our pilots of heavy aircraft to see if this is a reasonable interpretation of what we see there.
I was curious so I went googling, and one thing I did find was that the A320 and similar other aircraft (Boeing 767 for example) can be susceptible to oscillations during a crosswind, if they are on auto-pilot. This makes me wonder if the auto-pilot is to blame here.
A few things. First, there is no crosswind from the point of view of the plane, it’s just flying straight through the air. It’s only a crosswind for the aircraft once you start trying to put the wheels on the ground. It’s only then that the relative movement of the air over the ground has any relevance. With that in mind, an autopilot can’t be susceptible to oscillations in a crosswind because it doesn’t know there is a crosswind. It may be a factor in an auto land, but again that would only be at the point where the plane is getting lined up with the runway, ie a few seconds prior to touchdown. That said, I fly the A320 and am not aware of a tendency for the autopilot to have oscillations during a crosswind autoland.
Second. Those are not “oscillations” in the context of what an autopilot does, those are “wild gyrations”. An airliner would never be certified if the autopilot did that shit.
If I didn’t know better I would think it was deliberate wing waggling to say hello to the pilot’s mother on the ground or something, but that would be incredibly unprofessional, and highly unlikely.
Over controlling seems unlikely as I can’t imagine a pilot being that bad and the other pilot not taking over.
Turbulence is a possibility, but it goes on for a long time.
Edit: I see where you probably got the reference to autopilot oscillations from now. I think that poster on PPRuNe doesn’t know what they’re talking about for the reasons I gave above. However, knowing that it as an A320 leads more credence to the possibility of over controlling in turbulence. The A320 is a bit weird to fly because in the “Normal” flight control law, the flight computers try to maintain whatever attitude the pilot sets. If that attitude is perturbed by turbulence the flight computers will try to correct. If the pilot also tries to correct it can result in a bit of over controlling. That’s pretty extreme though.
The following incident may be relevant, though it occurred during landing rather than a go-around. I’ve quoted a bit about normal law and pilot induced oscillations.
Small spoilers (sometimes called spoilerons) can be used in conjunction with (or instead of) ailerons for roll control. Much smaller than the lift dumpers deployed on landing.
My first reaction was also a wing wag but the number of oscillations doesn’t make that seem likely.
I’ve almost been rolled in a small plane because the wind exerted more force than I could counter with ailerons. But never a succession of wind that caused gross over correction.
I can’t even guess what was going on in the video.
If it was a crosswind event then I would steer into the wind and climb out to minimize the effect but that’s in a small plane.
Yes. Spoilers are used for roll control as well as lift dumping / speed reduction.
One of the reasons it is not recommended to use “into wind aileron” during a crosswind take-off, a standard light plane technique, is that it also lifts the spoilers which adds drag and reduces the acceleration. This invalidates the take-off distance calculation.
His conclusion is that “rotor turbulence” coming off Gibraltar Rock combined with the always-entertaining Airbus autopilot system caused the oscillations.
A well presented analysis, and a cute dog. Unfortunately I think his opinion is worth as much as anyone else’s who wasn’t there. If there’s an investigation it may not go any further than the company safety department, though hopefully other A320 operators get some information.
I think he is saying the autopilot is reacting appropriately to the turbulence.
He is definitely talking about the autopilot, not the flight control laws. The laws are entirely separate from the autopilot and a pilot, even a Boeing one, would not confuse the terms.
Mentour Pilot actually did a video about this yesterday. You can watch it here: The "Rocky" Gibraltar Go-around - What happened?! - YouTube . If you don’t want to sit through the whole thing, he says it’s basically the autopilot reacting to “rotor turbulence” caused by Gibraltar’s unique geography (i.e. building an airport right next to a huge rock in an otherwise relatively flat area).
If you’ve not heard of his channel, he’s a professional airline captain (PIC) flying 737’s, and some sort of trainer/mentor for his company (don’t really know or understand that part of it). He’s normally rather reluctant to comment on stuff like this, but chose to do so this time for a variety of reasons he mentions in the video (biggest is that it had a happy ending and is explainable with objective facts).
