Caravan Pilots

Page 2 of 2

Author:  pdw [ Fri Mar 25, 2011 2:04 am ]
Post subject:  'Dead tired' Mayday Program (Discovery Channel)

Discovery-Channel's NTSB documentary summarizing Flight 3407's Feb12/09 accident sequence affirms "ice on wings" and that 'pilot "fatigue" ' is suspected. A "relay switch" for the earlier stall-warning, activated manually when opting for the higher Indicated-Airspeed(IAS)-trigger in icing, is found among debris and determined to be 'on' (confirms the low-IAS-warning mode for icing stall-speeds was in use). All OK prior to arriving at (slowing to) the desired / lower airspeed in final approach (where gear is just extended and flap-angle in process of increasing) despite ice build-up in gust-prone headwind (HDW), flat-approach, and in 'shallow turning' (heavier load factor ) to marker & glide slope. Gusty HWD / relativewind(RWD) with ice requires wary eyes on airspeed to catch rapid lift losses needing quick 'positive power adjustment' for countering unexpected IAS bleed-offs despite 'higher airspeed minimum' mode for safe icing airspeed-maintenance. Extra 'bleeding' during 'intentional speed-reduction for approach' slips by the pilot's field-of-vision to easily and is more precarious with airframe ice (can lose the 'safe airspeed' more quickly, unexpected). 'More-gradual / low-turbulence RWD-loss' leads to smooth/less-recognizeable loss of IAS in addition to added lift-loss dangers of a coarsening ice-catch. Sudden/unexpected increase in 'slowing rate' erupts when RWD reduces at a decreasing-side of a HWD-fluctuation [ the Decreased Performance Shear (DPS)]. The abrupt loss of RWD fast-erases a smaller 'margin of safe approach-speed' still remaining above an ice-catches' high stallspeed. DPS made IAS drop very-suddenly more-sharply (one reason why low speed might not be noticed soon enough) during a reduced/low power state of approach with ice.

NTSB narrative identifies 3407's accident as 'fatigue-related control-loss' which starts during/after 'slowed airspeed' with an unusual control-response/pitch-up with ice. While intentionally slowing for glide-slope the expedited slowing rate coincides with higher developing icing stall-speed and (likely pending determination by NTSB) a loss of concentration brought on by suspected "fatigue". 3407 abruptly steers 'excess nose high' (not clear why) just after stick shaker activates and autopilot switches off upon activation of low airspeed warning (emergency signal). The simultaneous steer-up action (NTSB 3407 Video) can be the result of 'feeling for backpressure' as the autopilot dumps control to pilots, the abrupt end to a stretched-out slowing period after the plane at-first appears 'stubborn to slow' (an airspeed-boost from updraft and/or HWD-increase). Video animation of flight recorder-info further shows a 'power addition' to be after 'up pitch'/'nose up', so evidently the 'abrupt attitude change' can't be related to engine-thrust surge of the forthcoming power-up, since power for recovery is not-yet levered at that point. We are informed about the 'flap retraction', which next is the 'close following' flap-withdrawal once 'full power' in place. 'Flap release' is portrayed there as hindrance to keeping control though it is not clear if 'flaps up' action alone prevents the airspeed recovery. It's difficult, to know for certain if IAS was sufficient during 'late raising flaps' or where control-loss begins (NTSB recorder info). Prior to extreme slowing, airspeed is nearing iced stallspeed where 'low IAS-warning' triggers, the place DPS steals remaining safe-margin so much faster. And the 'nose up' at the low power setting is almost certain to have wiped away anything that was left of that margin, seeing this 'safe airspeed' needs to ride so much higher above the 'iced stallspeed'.

A typical/intentional slowing period to decrease speed inside 'three-miles to marker' grows longer unexpected as strong / positive HWD fluctuation meets slipstream, which gives a boost of airspeed/lift despite appropriately-levered power-reduction. GS slowing-effects of ice-DPS will seem 'less' to pilots as GS-inertia is consumed, where a burst of increased RWD momentarily simulates 'more power' at the light engine-thrust setting (IAS leaps higher as HDW-rise consumes GS-momentum). In rapid succession comes the 'close following' DPS of lost RWD, the loss of HWD (NTSB Video / Wunderground evidence) bleeding-off airspeed/IAS into a precarious power/airspeed deficit very quickly and at the high (closer) stallspeed. Such negative-shear action is traceable here among multi-station WX history data (Wunderground/Madis 4 minute intervals) as downward fluctuation of nearby wind speed ('decreased performance' that trails a RWD-boost). The RWD fluctuation is to higher-RWD (Increased Performance Shear - IPS) then to much-lower RWD (DPS) while crew anticipates cessation of their intentional airspeed reduction. Initially the desired/anticipated slowing curve is with the stubborn IAS-decrease but switches (in a blink of an eye) into unexpected and expedited airspeed-loss, into the 'low power experience' within low groundspeed/inertia. The DPS shoots airspeed below 'crew intention' too quick (as the negative gust meets slipstream). The 'bleed' is harsh as RWD/IAS -cut (DPS) adds itself abruptly & without forewarning to slowing the airspeed faster so that the already-sinking airspeed plummets unexpectedly-rapid while now having less IAS to-lose in dropping down to a higher/'ice elevated' stallspeed-range lingering 'active' from prior enroute ice build-up.

An extra 'slowing factor' crops up as HWD 'lets up' in the middle of a routine airspeed-reduction in level approach at ~1500-1600'AGL (2300MSL). Left turn to 'marker heading' met 100% HWD-component (100%-incident to the gust-type DPS) so IAS safe-margin is very suddenly lost quick (highest rate of 'slowing factor'/RWD-loss per knot of shear strength). A light DPS shear-transition now gives its highest possible Decreased Performance for a more-intense slowing, which joins the ice-catch'es influence to make the slowing the worst it can be. The 'extra slowing' enters at the trailing end of the slow-down period and thus also catches crew off-guard for the hastened speed-of-arrival of their awaited ideal slow speed. The strong onshore breeze flows a bit wave-like via its rise/fall as it contours over varying AGL altitude, the uneven terrain; it is moist inversion air overriding KBUF's lake head elevations producing the mechanical turbulence (MECH) after rising from the flat Lake Erie's warmed surface after a previous day's 18 C [Erie PA. Wunderground history]. Brief delay to notice extra low-power/IAS of a shear-effect quickly turns to crucial delay for crew to lever any extra power soon-enough, resulting in a sudden state of overwhelming demand for thrust to recover with an ice's handicap. 'Safe airspeed margin' maintained above heightened stall-speed of ice then drained away in shorter time during the intentional slowing and including extra DPS-bleeding, far more than it usually takes when that IAS is above a normal/lower/clean stallspeed-airspeed. In this approach the strong-enough southwest HWD facilitates 'lee effect' where 'WSW' wind-directions slope into lowlands (MECH downdraft) at the low-alt co-ordinates of this accident sequence, which can also be related to sudden unexplained 'pitch up'.

Author:  pdw [ Thu Mar 29, 2012 10:58 am ]

The NTSB and Mayday program worked together well in explaining the details of the investigation.

Page 2 of 2 All times are UTC [ DST ]
Powered by phpBB © 2002, 2006 phpBB Group