Narrative:

Coming in on the arrival we were vectored off at least a dozen times. After going south of the airport then on an extended downwind; we were finally given a heading to intercept the localizer. After we began to capture the localizer the controllers were giving us some speed tolerances and at around 25+ miles our plane began a severe right turn breaking off the localizer. At about the same time it became obvious the approach controller asked where we were going. I pulled the heading knob and began a left turn back. Once established on a heading I reengaged the localizer; at the same time the controller came on and told us to maintain 160 kts and the localizer broke off again; this time towards the left. Again I switched to heading mode and with the winds at altitude being out of the southwest at around 45 knots it took a great deal to get back on course. A final attempt was made to engage the approach again but again this time it started a hard right turn; I disengaged the auto pilot and began flying the plane and using raw data to regain the localizer and GS. I believe somewhere outside of [the FAF] we were completely established and unfortunately had a scud layer at about 1000 ft to go through before making visual contact with the runway. The raw data was working but the auto pilot was not responding correctly; and unfortunately due to the issues at hand we didn't get the opportunity to try the other ap.because of the stress of the weather for the night and due to a long flight; combined with the constant vectoring and speed changes; I cannot completely say it was not an operator issue. I do know that any time the ap was engaged for the approach it vectored off and that when I hand flew it with the flight director it worked fine. Again I would like to have had the opportunity to try the other ap; I truly believe it was a mechanical malfunction. I'm not really sure how this could have been prevented beforehand but it certainly gets the attention in my book to kick off the auto pilot more often and just fly the plane.

Related reports:

Google
 

Original NASA ASRS Text

Title: MD-11 First Officer reported the autopilot failed to track the localizer on approach.

Narrative: Coming in on the arrival we were vectored off at least a dozen times. After going south of the airport then on an extended downwind; we were finally given a heading to intercept the LOC. After we began to capture the LOC the controllers were giving us some speed tolerances and at around 25+ miles our plane began a severe right turn breaking off the LOC. At about the same time it became obvious the approach controller asked where we were going. I pulled the heading knob and began a left turn back. Once established on a heading I reengaged the LOC; at the same time the controller came on and told us to maintain 160 kts and the LOC broke off again; this time towards the left. Again I switched to heading mode and with the winds at altitude being out of the southwest at around 45 knots it took a great deal to get back on course. A final attempt was made to engage the approach again but again this time it started a hard right turn; I disengaged the auto pilot and began flying the plane and using raw data to regain the LOC and GS. I believe somewhere outside of [the FAF] we were completely established and unfortunately had a scud layer at about 1000 ft to go through before making visual contact with the runway. The raw data was working but the auto pilot was not responding correctly; and unfortunately due to the issues at hand we didn't get the opportunity to try the other AP.Because of the stress of the weather for the night and due to a long flight; combined with the constant vectoring and speed changes; I cannot completely say it was not an operator issue. I do know that any time the AP was engaged for the approach it vectored off and that when I hand flew it with the flight director it worked fine. Again I would like to have had the opportunity to try the other AP; I truly believe it was a mechanical malfunction. I'm not really sure how this could have been prevented beforehand but it certainly gets the attention in my book to kick off the auto pilot more often and just fly the plane.

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.