Thick Investigators?
Can the DFDR discriminate between a pull force and a "back-trimmed in" autopilot auto-trim induced resultant force? The DFDR may have logged the nett force resulting from the pilot pushing against the yoke's nose-up out-of-trim force.
From the article circulating:
"The NTSB said a stall warning device known as a "stick shaker" appears to have behaved properly, activating when the plane's speed dropped to 130 knots (150 mph). At that point, however, "there was a 25-pound pull force on the control column," pulling the plane upward, and data suggests there was a "likely separation of the airflow over the wing" -- meaning the plane had stalled."
*
They don't seem to be able to get it into their heads that the aircraft had been already progressively (and totally) back-trimmed (during its unmonitored deceleration towards the stall) by the autopilot's elevator auto-trim into a zoom-climb condition. Add the pitch-up effect of max power and you have an aircraft PRIMED to pitch up suddenly. The pilot would have added max power and instantly faced having to fight, with forward yoke pressure, a massive nose-up pitching moment.
Forget tail-plane stall. Think more in terms of the threat posed by partial automation. An autopilot that could be set to level off from a descent - but no auto-throttle to add power. All it took was the pilots to become momentarily distracted and forget to add power to maintain the level speed.
*
Partial automation is a deadly trap i.e. an autopilot that levels off and auto-trims the elevator back all the way into a stall. It's very similar to the 737 crash in Amsterdam where the auto-throttle was supposed to add power - but their autothrottle had been disabled by the radar altimeter "spiking" to 8ft and thereby*replicating what normally happens in the landing flare at 10 feet above the bitumen. That radar altimeter flaw is a known fault by the way (and had even happened before on that same airframe).. The pilots in the 737 accident responded in a very similar manner i.e, max power once the stick-shaker/stick-pusher warning annunciated and alerted them. However they (he actually, because the other pilot wouldn't have realized what the problem was) would have then been fighting a huge rearward pressure on the yoke, caused by the full back elevator trim and the pitch-up effect at low speed of underwing-mounted engine thrust. In both accidents an unrecoverable low altitude stall resulted.
*
The only essential difference between the two accidents is that the Dash8 doesn't have an autothrottle...... but that quintessential and vital fact can be overlooked by distracted and fatigued pilots heads down re-programming the FMS keypad during an approach.
*
So it's pilot error very ably augmented by a phenomena destined to be known as partial (impartial?) automation (in the Dash8 case). In the 737's case it's the lack of an adequate deceleratory alerting system. The very rapid reduction in speed of a configured/configuring aircraft - once an autopilot captures a dialled-up height - trumps the alerting threshold value of a stick-shaker/stick-pusher. Why? Mainly because the stall recovery will be severely complicated by the pilot suddenly having to wrestle with a 25 lb zoom-climbing stick-force gradient.
*
In both accidents the pilots were unable to avoid a full-blown nose-high stall at low-level and the ensuing deadly autorotative wing-drop. It's possibly the deadliest surprise trap you could ever spring upon a pilot at low-level, particularly if he's still in IMC (or is flung back into cloud by the zoom-climb). It'd be totally disorienting. Is it going to cost Bombardier and Boeing? Good question. Past experience of such inbuilt technical quirks (eg Helios 737 and lack of attitude indicator twinning, center fuel-tank inerting, MD-11 flawed smoke checklist etc) would suggest that they will introduce a flight manual warning and stop short of any modifications. Why? Introducing a modification, even an additional warning system, would be a tacit admission of a pre-existing deficiency that would make them very vulnerable to litigation.
*
And so we are stuck once again in that round-robin of immoral rectitude that makes an expected call of "pilot error" - when it patently and obviously (to the cognoscenti)*really isn't.
*