AA757 Near Stall - Recovery Caused Injuries
 

"Crew failed to monitor speed loss before 757's aggressive manoeuvres

By David Kaminski-Morrow10 May 2022

US investigators have found that aggressive control inputs to increase poorly-monitored airspeed resulted in injuries to three flight attendants on an American Airlines Boeing 757-200 descending to New York JFK.

The incident occurred as the aircraft was arriving from Edinburgh on 6 September 2018.

According to the newly-released final conclusions from the US National Transportation Safety Board, the autopilot’s ‘altitude hold’ mode engaged as the 757 neared its pre-selected altitude of 12,000ft having descended from 24,000ft.

Some 16s later the aircraft’s altimeter setting changed – from the standard cruise setting of 1,013mb to 1,020mb – and this caused the altimeter immediately to display a higher altitude of 12,100ft.

The captain, who was flying, opted to re-acquire the desired altitude of 12,000ft by selecting ‘flight level change’ mode and the autothrottle responded to enable the aircraft to descend.

But the crew also chose to decelerate slightly, from around 255kt, by selecting an airspeed of 250kt – which caused retardation of both thrust levers. When the thrust levers reached their aft stop, the autothrottle entered ‘throttle hold’ mode.

The cockpit crew, which included the two pilots and an international relief officer on the jumpseat, undertook an approach briefing during which air traffic control cleared the jet to descend to 7,000ft.

Investigators found, however, that the aircraft – still in ‘flight level change’ mode – was trying to maintain its selected altitude of 12,000ft having reduced thrust. The airspeed started bleeding away and the 757 pitched nose-up to try to hold altitude.

The inquiry says the relief officer remarked three times in succession about the declining airspeed, over the space of 5s, and recorded data showed it had fallen to 186kt.

Thrust levers were suddenly advanced, the autothrottle exited ‘thrust hold’ mode, and the engines powered up.

“Push, push it over. Get the autopilot off, push it over,” the relief officer stated loudly, concerned that the airspeed was “dangerously low”.



Both the autopilot and autothrottle were disengaged and, with the aircraft 10° nose-up, the control column was moved quickly forward to command nose-down attitude.

“These aggressive manoeuvres caused the flight attendants in the aft galley to be thrown against the ceiling,” says the NTSB. One of them suffered a fractured arm, and two others were injured.

The pitch varied between 10° nose-up and 3° nose-down, with fluctuations between positive and negative g-forces over an 11s interval before the airspeed, pitch and control column positions stabilised and the jet started descending to 7,000ft.

It subsequently landed without further incident. None of the 104 passengers or the other six crew members was injured.

The captain had accumulated over 13,600h on type while the first officer had 565h.

Investigators state that the crew’s failure to monitor the airspeed adequately led to aggressive control inputs in an effort to accelerate, resulting in the injuries to the cabin crew."

From Flight International.

Gotta admit, I like to momentarily check the flight instruments while giving the briefing, just to ensure that there is still a sort of monitoring function.

Report DCA18LA285

Did we enter some sort of parallel universe a few years ago - one in which pilots don't know how to fly, or scan their instruments? :confused:

Is it usual to brief for NY when half way down the descent? We used to brief during the cruise and cover the options so we could add a mini-brief if the situation changed. Bugger all else to do when crossing the Atlantic

Even with a qualified relief pilot yelling in your ear? I suspect there were numerous nervous glances and interesting body language at each other once the ship was righted.

Wow, definitely a parallel universe when it takes almost 4 years to produce a report from an incident where the aircraft and all crew and pax survived!

It's just too easy to be a keyboard warrior. Thousands and thousands of safe flights everyday so I don't think it's appropriate to deduce from isolated incidents that pilots these days don't know how to scan and fly. Lessons to be learned for sure but no need to dramatize and generalize, I think.

Thousands and thousands of hours sitting behind pilots in the simulator from which I can tell you........ pilots these days don't know how to scan and fly. There is a need to dramatize and generalize. Think of the manifold and lethal ways in which gross ineptitude has been demonstrated at Air France for a start and then move eastward from there to where the true horrors begin.

Well, let's see. Off the top of my head there has been, in recent years :

A crew who also did not watch their airspeed and crashed short of the runway at SFO.
A crew who forgot to select go-around thrust when attempting to go around.
A crew who almost forgot to take off.
A crew who did take off but without any airspeed readout because the pitot covers were left on.
A crew who took off with both engine cowl flaps undone.
A crew who attempted an approach while so hot and high they forgot to lower the gear, scraped and damaged both engines along the runway, somehow got airborne again, only to crash.
A crew who after landing in a crosswind used the yoke like a car steering wheel to try to stay on the centre line, instead of the rudder, and went off the side of the runway.
A crew who landed so deep and in a tailwind, they went off the end of the runway.
A pilot who held full back-stick at cruising altitude, fatally stalling the aircraft.
A pilot who broke off the fin by incorrect use of the rudder during turbulence

Were these not all very basic piloting errors ? Seems like a parallel universe to me !

PS, Am a 'joe average' qualified long haul and short haul pilot, not an armchair amateur :ok:

Didn't mean to call you an amateur. But we are all Monday morning quarterbacks here.

Your list above is valid and relevant but when put into context of more than 100.000 commercial flights a day around the world, that's 36.5 million flights in just one year, I think we must be doing something right. The AA brand alone has more than 6000 daily departures.

I am wondering about the parallel universe where 186 kts is a dangerously low airspeed requiring aggressive intervention

I am wondering about the parallel universe where the autopilot remains in FLCH mode when the aircraft levels off at the selected altitude. Is there something different about the B757 to other Boeing types (eg B744, B777)?

Nope. Approaching the MCP altitude it should go to ALT CAP then SPD/ALT HOLD. The Auto throttle should wake up.

Mode confusion in some iron aircraft? Think about if this had happened in some bus?

It would had been the P2F south-east Asia cadet's fault

Good examples

But are we going to pretend similar accidents didn't happen in the past?

None taken. I agree that on the whole flying is very safe, but I am getting increasingly alarmed that some very very basic mistakes are being made. These are not caused by complicated equipment failures - or any equipment failures - these are caused by pilots not being pilots; not looking at their instruments, not doing a proper walk around, not controlling the automatics etc.

These basic mistakes should not be happening - there are at least two pilots on these flights and the SOPs have a large amount of cross checking and confirmation between the pilots to catch all the small errors we all make. And by small errors, I do not mean failing to scan the instruments; that is fundamental. And if the automatics are not doing what you need, the pilots should catch this and take appropriate action. Jet pilot 101.
The pilots in this latest incident not only failed to monitor the automatics or watch their speed, but it seems they panicked in their recovery, instead of smoothly reacting.

Granted, we might not all be as successful as Captain Sullenberger and F/O Stiles were when faced with a double engine failure at low level - fantastic piloting - but we should at the very least be capable of flying normal manoeuvres and scanning correctly.

When I fly and even when I brief, I regularly glance at my PFD - just as you do when driving your car, you (should) regularly glance in all three mirrors as well as looking forwards so you know what traffic is around you, as well as monitoring your car speedometer. Same with aircraft instrument scans.

I seriously think that training, testing and recurrent Sims need a big shake up. Really the XAAs should conduct them so there is no favoritism or knowledge of the pilots.

edit to add answer to SpamCanDriver, No, but in aviation we are supposed to learn from everyone's previous mistakes and not repeat them. We seem to be going backwards safety-wise at the moment.
.

Well, absolutely nothing exciting would have happened because the Airbus doesn't have the moronic HOLD mode fitted to the Boeing auto-thrust system. It would simply have changed from THR IDLE to SPEED.

HOLD is a hangover from the previous century when it was necessary to physically remove power to the autothrottle servos to prevent obscure problems instead of fixing those issues in the first place.

I disagree that HOLD mode is moronic. What is moronic is the mis-handling or lack of awareness and understanding of the aeroplane systems operation.

On the B777 and B787 the HOLD mode software has now changed in that speed protection, i.e. Autothrottle Automatic Activation, now functions when in that mode. I presume the change applies to other models as well.

Absolutely agree we should

I was just referring to the theme that pilots wouldn't of made these basic flying errors in the past, when it was a more manual job.
They did make the same errors