HarryMann;
How much manual handling sim-time and real flight-time is built up per thousand hours in heavy to severe turbulence conditions (i.e. when it would perhaps be expected the a/p would find itself on or beyond limts) ?
A thousand hours is, for most major carrier transport pilots, about 12 to 18 months of "hard time" depending upon contracts. (Putting in 85hrs of hard time in a month for an airline pilot is risking high fatigue levels). A 330 captain on long haul operations may get 3 landings a month, perhaps 4, maybe 5 on rare occasions. Domestic crews will do 75 to 90 landings a month and about the same hard time.
From my own experience, almost all crews engage the autopilot above 3000ft. Very few fly to 10,000ft/FL100 and almost nobody flies to cruise altitude. Disconnection can occur at about the same altitude on approach - 3000ft, but most crews disconnect below 1000ft, and many at around 400ft.
At 1500ft on approach, an airliner is just over two minutes to touchdown.
Out of a typical flight, I would estimate the first five minutes and the last 3 minutes are hand-flown, the usual exceptions applying.
If the typical number of landings applies per month, a long-haul crew member may be getting, at the most, 30 minutes of hand-flying per month or about 6, perhaps 9 hours of hand-flying during the 12 to 18 month period. A domestic pilot would be doing the number of landings per month times about 7 minutes but it's more difficult to estimate for short-haul.
There is NO hand-flying taught or encouraged in the simulator except for the requisite steep turns and approaches to stall on an initial aircraft checkout. There is NO standby instrument practise and, very rarely, we did a raw data, (no flight directors) hand-flown, (including manual thrust levers), ILS. These approaches are usually between a bit ropey and a go-around but some are spectacularly, beautifully hand-flown with the loc/g-s or airspeed "painted" on the gauge face. Over the years, hand-flying was discouraged more and more and finally the authority to disconnect was removed from the FCOM except in circumstances where workload was low, (no traffic, no weather).
I have never in 35 years of sim work, carried an airliner into the stall. Such work was and should be test-pilot work as no airline pilot ever expected to have to deal with a full stall and certainly not a spin in a heavy transport.
With four, likely five recent accidents in which the crew lost control and stalled their aircraft resulting in a fatal accident, perhaps times and expectations and reasons for studying the full stall, have changed.
Within limits, we never had altitude restrictions within which we had to recover and we always watched for the secondary stall from pulling to hard or not enough nose-down. Jet Upset maneuvers were never done nor trained. We practised EGPWS and TCAS escape maneuvers.
I don't know how much of this reflects others' experience.