PPRuNe Forums - View Single Post - Continental TurboProp crash inbound for Buffalo
Old 20th Feb 2009, 20:19
  #766 (permalink)  
Flight Safety
 
Join Date: Jan 2001
Location: Dallas, TX USA
Posts: 739
Likes: 0
Received 0 Likes on 0 Posts
From looking at these 4 NTSB reports (which I posted earlier), here's another possibility for what happened to flight 3407. I seem to remember someone else may have speculated about this earlier.

LAX06IA076
DCA01MA031
DCA97MA017
DCA93IA027

If the autopilot was set to ALT HLD (for 2300 ft), AND the airframe and/or wings were accumulating ice, it could have happened like this.

As the aircraft descended and slowed, then leveled at ALT HLD at 2300ft, ice buildup could have added to the airframe drag, and loss of lift from the wing. Then as the pilots made power adjustments (to around 134kts +/-) for flaps 5, then the landing gear, then flaps 15, the autopilot could have been adjusting the stabilizer trim to a higher AOA (more nose up) to compensate for the airframe drag and loss of lift at that airspeed, to maintain 2300ft.

As the flaps were transitting to 10 degrees, the combination of airframe drag, loss of lift and airspeed, along with the increased AOA necessary to maintain 2300ft, may have triggered the stick shaker and pusher, which disconnected the autopilot. At that point, the stabilizer trim may have been adjusted significantly nose up, compensating for the drag and lift loss to maintain that altitude. This might explain the immediate 31 degree nose pitch up, after the autopilot disconnected. And if there was ice accumulated on the wings (beyond the boots) and on the airframe, it would have made the stall worse, and recovery that much more difficult.

However to counter this theory, the NTSB has said they did not hear significant trim motor sounds on the CVR. That said, the FDR would probably give a more complete picture of this.

Last edited by Flight Safety; 20th Feb 2009 at 20:30. Reason: To make the language more clear.
Flight Safety is offline