News:

Precision Simulator update 10.173 (24 February 2024) is now available.
Navburo update 13 (23 November 2022) is now available.
NG FMC and More is released.

Main Menu

Autopilot Disengagement on Manual Override

Started by emerydc8, Thu, 14 Nov 2019 18:20

emerydc8

According to an incident report on a JAL aircraft years ago, the 747-400D A/P will not disengage if you manually override it. Does anyone have any information on whether other models will disengage? I've tried to get the 767 A/P to disengage by overriding the aileron, but couldn't do it. There is nothing in our manuals about it.

QuoteManual Override of Autopilot
"Manual override" means that when the autopilot is engaged, a force applied by the pilot to the control column makes the control surface angle different from that of the autopilot command. The term "manual override" is used with this meaning in the remainder of this report. With the Boeing 747-400D airplane, when only one autopilot is engaged, the pilot can manually override the autopilot by applying a force on the control column. The autopilot does not disengage due to the manual override. If the pilot applies a force to the control column and manually overrides the autopilot when only one autopilot is engaged, the FCC monitors the autopilot servo positions and elevator angles issues an autopilot caution if they disagree for more than six seconds (amber "AUTOPILOT" caution message displayed on the EICAS (Engine Indication and Crew Alerting System) display, aural warning, and illumination of the master caution light).

Will

I can't speak to what happens in Boeing aircraft, but inadvertent (and unnoticed) autopilot disengagement via manual override is what brought down Aeroflot 593 back in the day. That was an A310.
Will /Chicago /USA

Roddez

Not to mention Eastern Air Lines Flight 401 back in 1972...

The crash occurred while the entire cockpit crew was preoccupied with a burnt-out landing gear indicator light. They failed to notice that the autopilot had inadvertently been disconnected after a crew member leaned against the control column.  As a result, the aircraft gradually lost altitude and crashed.

Rod.
Rodney Redwin
YSSY
www.simulatorsolutions.com.au

John H Watson

This stuff is hard to make sense of. The tech manual text is bogged down with technobabble.

It seems that manual override generates what is known as loss of detent or a camout condition. This is where the autopilot servo and control surface positions do not agree.

Here's what my book says:

General
The hardware monitor logic gets arm and engage signals from
software logic.... If conditions are proper, this
logic energizes the arm and engage relays. The hardware logic
required for arm and engage is:
• Not disengaged from wheel switches
• Power supply valid
• Computer valid
• Not power-up test inhibit
• Detent logic
• Camout logic


The power-up test inhibit makes sure that the autopilot can not
engage during an FCC power-up.

Detent Logic
The detent logic compares the servo position and the surface position
from the autopilot servo LVDTs. A detent condition exists if the
servo position is equal to the surface position for all three servos
(elevator, aileron and rudder). A loss of the detent condition from
this logic causes an autopilot disengage only if the autopilot is
multi-channel engaged.
If single channel engaged, this logic causes
an autopilot caution annunciation.

Camout Logic
The camout logic in each FCC monitors the detent status and engage
status of the other FCCs. Relative left or right channel camout causes
the local channel to disengage if the foreign channel disengages with
a detent trip and if it is the only other channel engaged. In other
words, for dual channel engaged, if one channel has a detent trip,
both channels disengage. If triple channel engaged, a detent trip
disengages only the channel which has the detent trip.
Non-Isolated Disconnect
During dual channel engagement, a detected failure of two same
inner loop or autoland sensors causes a hardware disengagement.


Single A/P detent trip/camout only generates cautions.
Manual override during multi-engage would cause detent trips for all three systems. I'd say this would result in disengagement in dual and triple (engage).

Hardy Heinlin

I learned during my first Lufthansa 744 sim sessions many years ago that the A/P can be disengaged by a manual counterforce of about 20 or 40 kg (elevator axis). I can't remember whether we actually tried it, but the feature was taught at LH anyway.

Maybe this "40 kg" condition is a second logic layer on top of the logic described above. The required force is so great; the pilot cannot apply it accidentally.

emerydc8

So according to John's technobabble, during single autopilot operation, a manual override will not cause a trip? It has to be multi-channel for a trip? Maybe I'm reading too much into the JAL incident report, but it sounds like other models of the 744 (not the -400D) could actually disengage on manual override - like Hardy said with the LH sim.

Hardy Heinlin

When the pilot is moving the stick to angle X, will the A/P servo mechanism also move to angle X?

If so, the relevant disagreement might be a disagreement between World A and World B:

World A: The stick and servo angle inside the aircraft (pilot and autopilot actions)

World B: The surface angle outside the aircraft (commanded by World A)


John H Watson

It's hard to describe with words. There are plenty of pictures in the technical manuals. With the a/p engaged, the autopilot actuator pistons engage with the output crank of the actuator.
Everything is then connected together:
autopilot servo, controls (yoke, etc), flight control surfaces.

If you apply a mechanical input, this fights the springs in the engage mechanism and this is sensed by the autopilot actuator input/output LVDTs.

Hardy, ref Elec Book 15 page 48/49.