Lyman,
On takeoff, there was a crosswind from the right at 19 knots, gusting to 28 knots. It lifted off in a slight left-wing down attitude. A right-wing aileron and subsequent rudder application brought the wings to level and the plane on proper heading in the crosswind. However, with the loss of the engine and pylon structure, the aircraft cg shifted aft 2%. With all slats extended the stall speed was 124 KIAS . The asymmetric slat retracted left wing stall speed was 159 KIAS, that being the speed in which directional control could be maintained. At KISA lower than 159, a stall warning shortly after liftoff would have occurred had it been electrically powered.
In a SIM setup to duplicate the conditions of failure, thirteen pilots did 70 takeoffs and 2 landings. They had been briefed, of course, on the flight ahead of their SIM exercise. When the pilots exactly duplicated the flight controls and pitch attitudes of AA 191, the aircraft was lost. However, most pilots recognized that the start of the roll left was actually a stall and lowered the nose to gain speed, recovered and continued flight. Landings were not a problem. 80% right rudder and 70% right-wing down aileron and nose down from the 14º climb attitude was required for successful recovery.
At the altitude they were at (about 300 feet), cutting back the power on the two operating engines would have been not the thing to do, they needed more forward speed.
As for the pylon maintenance, the (to be AA 191) aircraft was in for a service bulletin to replace the aft bulkhead spherical bearing. The pylons were suppose to be removed in accordance with the procedures defined in the DC-10 Maintenance Manual (engine first, then the pylon from the wing). However, American had been experimenting with the use of a forklift to remove the engine and pylon from the wing at the same time to avoid a number of electric and hydraulic disconnects. American contacted McDonnell Douglas regarding this procedure and was advised against it due to potential damage to the wing/pylon mounts. They had performed this one step operation on several (4) foreign DC-10 they were servicing. On the aircraft in question (AA 191's), they started the removal process towards the end of a work shift and left the forklift running for the next shift to finish the removal procedure. When that crew came in they had trouble removing the forward bulkhead bearings and then notice the whole engine stand was mis-aligned. The stand was then moved forward about 12 inches to enable removal of the forward bulkhead bearings. Sadly, the whole procedure was done out of sequence from what it should have been. Because of this, the aft upper flange was damaged, not only on this aircraft, but also at other US airlines which used a similar forklift procedures on 76 engine/pylons combinations. I think fatigue cracks were found on 17 engine/pylons. Damage to the aft upper flange lead to fatigue cracks being initiated. The damage was not reported because it was thought the damage was to a low stress component and was a minor repair situation at best as defined by the FAA at that time.