PDA

View Full Version : Accidents caused by failure of GA autothrottle modes


A37575
27th Dec 2009, 12:35
The recently published NTSB report on the One-To-Go MD80 accident at Phuket, mentions the crew failed to push open the thrust levers as part of the go-around procedure. The aircraft subsequently failed to climb. They apparently relied on the GA mode to actuate the thrust levers while failing to see it was not correctly armed.

The report stated the autothrottles move up at eight degrees of angle per second; meaning it takes the throttles 5.2 seconds to reach full thrust position. It further stated the engines on the MD80 are capable of accelerating to go-around thrust faster than the rate given by the autothrottles.

Does anyone have the equivalent numbers above for the CFM 56 engines of Boeing 737's? Ie How many degrees of thrust lever movement per second on GA with autothrottles?

In the simulator, we often see pilots pressing the TOGA button to initiate an autothrottle go around and if for some reason TOGA did not work, the element of surprise sometimes sees the pilot fail to advance the thrust levers manually for a critical few seconds. Like the crashed MD80 they failed to advance the throttles manually when TOGA did not operate and instead relied implicitly on the autothrottles to actuate. Having pitched the aircraft manually to a go-around climb body angle there can be an alarming speed decay before the pilot realises something is amiss and he gets around to manually shoving the throttles to max thrust.

Is there any technical reason why in a go-around the pilot should not first manually open the throttles rapidly to go-around thrust and once there, then press TOGA? After all, this only takes a couple of seconds and is a good insurance against TOGA failure or pilot error. It would have prevented the MD 80 accident mentioned earlier. With a planned rate of rotation of three degrees per second on the way to the go-around attitude of around 15 degrees nose up, the slight delay involved before the TOGA button is actuated and sets the the flight director to GA mode, should not prove an unsurmountable problem.

By the time the thrust levers had manually reached go-around thrust lever angle, the nose up attitude would not yet have reached anywhere near 15 degrees.

HundredPercentPlease
27th Dec 2009, 12:45
Could it be because you would then have one procedure for hand flown G/A and you would have to have another for a G/A off a fully coupled approach? I'm not sure your method would have the right consequences at 50' on a low vis approach.

I much prefer the Airbus solution where you just whack the levers forward as fast as you like. If you have flaps out then as soon as the levers hit the end stop then TOGA mode is engaged.

GlueBall
27th Dec 2009, 14:08
The accident is not caused by failure of A/T TOGA mode; it's caused by the PF not shoving the thrust levers forward when the A/T didn't or was too slow. Manual inputs are elementary and instinctive anytime the automatics are not performing as they should. During any approach, in any airplane, irrespective of A/T operation, whether autoland or not, the PF must always keep one hand on the thrust levers.

A37575
29th Dec 2009, 08:15
so he can keep the PITCH moderated to SEE the THREAT ahead),

Threat? What "threat?" A normal go-around is not a threat. Or has this whole Threat and Error management gimmick got pilots shooting at shadows?