Not exactly an airliner, but an exec twin that failed to make the UK register - the Mitsubishi MU-2. I was investigating stalling speeds in various configs after some lower than scheduled speeds were found on a C of A airtest, and carried out several airtests after ground checks and engine / airframe adjustments. Finally everything seemingly possible had been checked and a/c should have been perfect. On the next airtest the aircraft auto-rotated at the stall (power off & clean) despite immediate full opposite rudder. The spin developed fully and showed no sign of recovery using 'standard' techniques. There wasn't too much time available as entry to the 'stall' had been 5000ft amsl (normally sufficient as recovery usually only lost a few hundred feet at most) and ROD was off the clock estimated 6000fpm.
I could write a chapter on the next 20 seconds but I'll be brief. After trying in/out-spin aileron (spoilers on MU-2) unsuccessfully, time was running out fast and, in desperation almost, I smoothly increased to max power on the in-spin engine. After about 5 seconds of oscillatory spin the a/c flicked into a spin in the opposite direction which I arrested after about 1/2 to 3/4 of a turn. Resumed symmetric power after closing both levers and recovered at 1000ft above the sea.
Not an experience I'd like to repeat - fortunately I was pretty recent on spinning, having completed 2 years with the Rothman's Team about a year before, my specialist solo manoeuvre being the lomcevak.
Training spins are usually entered by applying pro-spin rudder from level flight (or from manoeuvre) - this one went unprovoked and with corrective inputs. Later extensive investigation revealed an engine problem that caused a slow insidious loss of power on one power plant with no cockpit indications. It would take too long to explain the MU-2's controls and aerodynamics, but the effect of this power loss was, on a 'trimmed' a/c, to reduce the stalling speed by about 7kts and set up all the conditions required for a spin.
Lessons? 1. The UK CAA were absolutely right in not certifying the type and 2. I was wrong, with hindsight, to fly to the actual stall (less than scheduled) - better to stop the exercise at Vs and report 'no stall'. 3. Not a good idea to have a fleet of 1 odd-ball aircraft. If there'd been another similar type in West Africa we could have compared flight characteristics.
rts