B737 autothrottle failure

[downwind]

Based on the link below;

https://en.wikipedia.org/wiki/China_Southern_Airlines_Flight_3943

“the captain attempted to level off the plane by raising the nose. The plane's autothrottle was engaged for descent, but the crew did not notice that the number 2 power lever was at idle. This led to an asymmetrical power condition.”

How often has this happened on the B737 family? (is a clutch malfunction rare, I realize the Turkish airlines accidents in Amsterdam was based on a faulty rad alt

This accident seems similar to the Taroom A310 one… https://en.wikipedia.org/wiki/TAROM_Flight_371

Can this malfunction happen on the never generation of NG/MAX models?

[sonicbum]

Can this malfunction happen on the never generation of NG/MAX models?

Yes, but the A/T will be automatically disengaged due to the thrust liver split and the red A/T will start flashing. According to the different MSNs there are different conditions, but all of them include a significant thrust difference along with control wheel input of 10 degrees or more.

[Tee Emm]

How often has this happened on the B737 family?
Have certainly witnessed this type of event in the 737 Classic simulator. During re-current training the instructor failed the clutch motor on the No 1 thrust lever during a clean descent on a DME arc at idle power. The autopilot and autothrottle were engaged at the time. As the aircraft decelerated in level flight and extended gear and flaps just before ILS glide path intercepted, the No 2 engine increased thrust to almost 80% N1 while the No 1 thrust lever stayed at idle. The aircraft was by then coupled to the ILS.

The control wheel was 45 degrees off centre but the crew seemed oblivious to the increasingly dangerous situation. When the simulator instructor dropped the hint that 45 degrees wheel angle coupled with huge throttle split was unusual, the captain muttered a exclamation and called for the engine failure and shut down checklist for the "failed" engine. At the point the automatic pilot gave up the ghost and the aircraft rolled into a steep spiral dive left and reached 60 degrees angle of bank and ten degrees nose down before the simulator instructor "froze" the simulator to discuss that situation.
Neither pilot did anything to correct the situation and the captain kept calling for the QRH engine failure checklist but without taking any action to prevent the aircraft rolling into the ground. There was nothing wrong with the engine and all he needed to do was restore equal power by pushing No 1 thrust lever open since it was only the clutch motor that wasn't working.
The airline was mainland Chinese. Indications were that ethnic culture was probably involved as the co-pilot watched the whole drama evolve without a word. . .

[vapilot2004]

Outside of the mechanical failures that can befall any automatic system eventually, the Classics had two common problems that would cause this behavior: the failure of a clutch switch (really the override sense) in the throttle assembly, and an inadequate A/T computer.

Smiths redesigned the A/T box, including the software, and these changes were finally adopted into an approved AD with solution sometime in the early new millennium.

[parkfell]

Perhaps one answer to cope with a (subtle) A/T fail, is to guard the TLs during their scheduled movements.
A critical time would be on a G/A once TOGA is pressed......
Guard the controls and TLs during the descent below MSA/2500'RA.....

[Switchbait]

The auto throttle is a convenience, it is not necessary to fly the airplane.

The sooner people think of it this way, the sooner the problems listed above will go away.

Be a pilot, fly the airplane, all of it, all of the time.

[RAT 5]

Perhaps one answer to cope with a (subtle) A/T fail, is to guard the TLs during their scheduled movements.

Indeed, but in todays 'rely on the automatics' common culture unless it is an SOP, or enshrined in the training future, it wont happen. Sadly. In my outfits I tried to encourage it on the line; little effect as many blank stares. "It ain't an SOP." Well, following through the TL's after TOGA on takeoff is an SOP. Checking GA thrust is set, no follow through except AT TOGA, is an SOP. Following through TL's on approach even with AT ON, SOP. Following through at thrust reductions reaching crz, or thrust increase for a VNAV climb is not an SOP. Why only some of the time and not all the time?
Guys have no problem accepting their laptop goes AWOL occasionally, but they seem to have blind faith that a/c computers never fail. Yet they still ask the questions about why the FMC seems to be giving unrealistic vertical path information. That is also a computer. They accept sensors in their car fail, so why not AT sensors? If it's not taught it not happen, yet we've seen recent GA crashes due to incorrect AT response and lack of crew awareness.

[parkfell]

The answer must be that the technique of 'following through' is taught during the MCC/APS course.

As threat & error management continues to be instilled during this initial multi crew training, I would suggest that any provider worth its salt would "encourage" it from day one. It is always important to fail the A/T during the MCC to prove the point. They become so dependent so soon. A rapid addiction occurs.

I could go down the route and say it is good AIRMANSHIP, not to be confused with good old fashion common sense.
No doubt the THOUGHT POLICE will be on my case any time soon.

[Uplinker]

It’s interesting, we get a lot of the “don’t rely on the automatics”, comments, which I agree with.

However, I have just converted onto the 737 after flying the Airbus 320/330, and I have to say, the autothrottle and automatic flight system of the 737 classic is awful, compared to the Airbus FBW. So I wonder how many of the “don’t rely” comments are really saying, ‘this particular aircraft type has very marginal automatics, so you have to watch them like a hawk’?

Any pilot should obviously monitor what the aircraft autopilot and autothrust/throttle are doing at all times - hopefully that goes without saying. But some automatic systems are much poorer than others in their operation and interface with the human pilot.

For example, pilots who are used to moving thrust levers often say that having non moving thrust levers is a bad thing, yet we see these problems with failed clutches or misunderstood autothrottle modes that I have never seen in 12 years of Airbus FBW flying.

[Goldenrivett]

yet we see these problems with failed clutches or misunderstood autothrottle modes that I have never seen in 12 years of Airbus FBW flying.
What a sheltered life you must lead.
Air France flight AF1620 - Aviation Accidents Database (http://www.aviation-accidents.net/report-download.php?id=210)

http://www.aviation-accidents.net/report-download.php?id=210