Narrative:

Cockpit and flight attendants started our duty day and boarded our aircraft which was a tech stop from ZZZZ1. Upon arrival; the aircraft was loaded and being attended by the deplaning fas. [The aircraft] was readied and departed with an uneventful take-off with a cockpit crew of three and an observer flight mechanic in the cockpit. About 600 feet the right autopilot was engaged. At 1500 feet; the aircraft began a slow pitch over resulting in a 300 FPM descent recovering to a shallow climb but returning to a descent. While in a turn; the autopilot disengaged itself with the flight directors continued erratic behavior. The flying first officer (first officer) was commanded to hand fly the aircraft as he climbed through 12;000 feet. The first officer commanded the right autopilot again and the flight path became erratic in the pitch axis; changing 2.5 degrees up and down while being closely monitored. The first officer took over manually; while we trouble shot and noticed the trim was out of the green band and in the nose down zone. The left autopilot was engaged and the pitch attitude and airspeed was just as erratic as previously while climbing out. The aircraft was hand flown to our cruise altitude of FL310. The left autopilot was again reengaged with acceptable performance until top of descent where it again became unstable and was disconnected. Flight continued with an uneventful hand flown approach and landing however; the low speed stability trim made handling challenging.suspecting a weight and balance issue; the purser was summoned to the cockpit at block-in and a quick discussion about our load determined that it was very possibly the passenger had changed seat positions while they sat waiting. After arrival in ZZZZ; a PA announcement was made about changing seats/sections and the customers were made aware of the problems it could create. A new load sheet with the probable seat positions of the flight in question was created and in fact; the tow mac (mean aerodynamic chord) changed from 23.8% to 21.7%. After shutdown I consulted with the pilot on duty who brainstormed with others including a chief pilot and a lengthy logbook discrepancy was entered.

Google
 

Original NASA ASRS Text

Title: Captain of a widebody aircraft reported problems with the autopilot for the duration of the flight.

Narrative: Cockpit and flight attendants started our duty day and boarded our aircraft which was a tech stop from ZZZZ1. Upon arrival; the aircraft was loaded and being attended by the deplaning FAs. [The aircraft] was readied and departed with an uneventful take-off with a cockpit crew of three and an observer flight mechanic in the cockpit. About 600 feet the right autopilot was engaged. At 1500 feet; the aircraft began a slow pitch over resulting in a 300 FPM descent recovering to a shallow climb but returning to a descent. While in a turn; the autopilot disengaged itself with the flight directors continued erratic behavior. The flying FO (First Officer) was commanded to hand fly the aircraft as he climbed through 12;000 feet. The FO commanded the right autopilot again and the flight path became erratic in the pitch axis; changing 2.5 degrees up and down while being closely monitored. The FO took over manually; while we trouble shot and noticed the trim was out of the green band and in the nose down zone. The left autopilot was engaged and the pitch attitude and airspeed was just as erratic as previously while climbing out. The aircraft was hand flown to our cruise altitude of FL310. The left autopilot was again reengaged with acceptable performance until top of descent where it again became unstable and was disconnected. Flight continued with an uneventful hand flown approach and landing however; the low speed stability trim made handling challenging.Suspecting a weight and balance issue; the Purser was summoned to the cockpit at block-in and a quick discussion about our load determined that it was very possibly the passenger had changed seat positions while they sat waiting. After arrival in ZZZZ; a PA announcement was made about changing seats/sections and the customers were made aware of the problems it could create. A new load sheet with the probable seat positions of the flight in question was created and in fact; the TOW MAC (Mean Aerodynamic Chord) changed from 23.8% to 21.7%. After shutdown I consulted with the pilot on duty who brainstormed with others including a Chief Pilot and a lengthy logbook discrepancy was entered.

Data retrieved from NASA's ASRS site and automatically converted to unabbreviated mixed upper/lowercase text. This report is for informational purposes with no guarantee of accuracy. See NASA's ASRS site for official report.