Challenger’s Stabilizer Wing Prone to Icing Stall
Coming so soon after NBC Sports Chairman Dick Ebersol’s Nov. 28, 2004, crash in snowy conditions in Montrose, Colo., and the Nov. 21, 2004, icy crash of the airliner variant in Baotou, China, the crash on Feb. 2 of another Challenger in similar circumstances should give the Federal Aviation Administration (FAA) pause for thought about the aircraft’s airworthiness. It’s barely credible that any Challenger pilot would nowadays be unaware of the concern about any amount of upper wing surface icing and its effect upon slow-speed, high Angle-of-Attack lift generation on take-off rotation (ASW, Dec. 6, 2004). So is there perhaps more to it? In this accident, despite having a near full load, it would appear that the aircraft didn’t leave the ground on Teterboro’s 6,013-foot-long Runway 6 in calm but snowy conditions. The pilots’ explanation is that they rejected the takeoff because they were unable to get sufficient back-stick to rotate and then were unable to stop on the contaminated runway remaining.
In fact, because of heavy rubber skid-marks, the aircraft coming to rest with its speedbrakes and reversers deployed and cockpit voice recorder (CVR) evidence, that was obviously the case. So the abort was clearly precipitated by the crew’s inability to rotate — and that’s also the interpretation of comments made at the scene by the surviving pilots. “As they tried to [get] airborne they couldn’t, they just lost control and they went straight through,” eyewitness Robert Sosa said the pilot told him.
The Challenger icing-related crashes to date have typically been associated with either the inability to rotate/lift off cleanly (Montrose, Teterboro) or the sudden, once- airborne, snap-roll due to the one-sided stall of dissimilarly iced wings (Birmingham and Baotou).
Early morning wind and sun-shadow is thought to have played a part in the Birmingham accident. It could have caused overnight accumulations of ice to be residually concentrated on one wing only, by the time N90AG departed at 1207L. If the wing not in sun or wind shadow was relatively clear of ice, logically the high-mounted T-tail would be entirely “not in shadow.” That might explain why the Birmingham CL600 and CRJ200LR Baotou jet were able to rotate and get briefly airborne. In the latest Teterboro accident, it may be that the pilots conscientiously scoured the upper wing surfaces by feel for icing contamination — but it would have taken a ladder to closely examine the top surface of the high-mounted T-tail for a snow and ice build-up in the below freezing conditions. The National Transportation Safety Board (NTSB) says the plane was not de-iced. That decision may have been based upon the pilots’ own observations of clean wings or their own brushing of any contamination off the wings (but not off the high-mounted T-tail; tails got no mention in the NTSB’s Dec. 29, 2004 cautionary alert). To view that document, see: http://www.ntsb.gov/pressrel/2004/041229.htm
Tail and wing airfoil sections tend to be from the same stable and have similar design characteristics (even though the horizontal stabilizer may be an inverted profile). The wing and horizontal stabilizer’s coefficient of lift once contaminated may both be subject to the same super-critical leading-edge laminar flow breakdown.
There is a lot to be said for having leading-edge devices such as slats and Krueger Flaps. They may add weight and complexity but they “take the edge off” the cruciality of the laminar flow required by the super-critical wing sections found on most modern jets. They also make contaminated runway takeoffs that much safer and consequently, slip-sliding ice-rink aborts for poor acceleration and handling less likely.
It is also instructive to review the report on the icing implications of the Dec. 16, 1997, crash of a Challenger CL600-2B19 airliner at Fredericton, Canada (see http://www.iasa.com.au/fred.htm).
Icing would appear to be a lethal Achilles Heel for this breed of jet. The NTSB ruled that a similar type of “clean” wing was partly to blame in the March 22, 1992, crash of a USAir Fokker F-28 commuter jet at New York’s LaGuardia Airport. Jets with such wings “have been involved in a disproportionate number of takeoff accidents involving ice,” the NTSB wrote in its findings. However on Dec. 16, 2003, another Challenger aborted a takeoff on Runway 24 at Teterboro and ended up in the mud with minor damage only. That accident was blamed on a weight and balance miscalculation.
There is a further possibility. In the CL600 you taxi, take off and land with the yaw-damper (YD) engaged. It is not unknown for the YD to kick off during taxi or on the takeoff roll. It is an idiosyncrasy of the way in which the YD perceives a ground-transmitted bump. It doesn’t do this once airborne, so it’s never been “a biggy.” The Master Caution flashes and the Flt Ctrl light on the 8/10 overhead panel will illuminate. Instinctively, the PNF (pilot not flying) will reach down and re-engage the YD, killing the Master Caution light. But Murphy’s Law has the YD and Autopilot (A/P) buttons alongside each other and they are the same size, identically lit, etc. If the YD kicked off on takeoff, it’s possible that the PNF mistakenly hit the A/P switch. With the heading bug aligned on runway heading, the airplane would’ve tracked cleanly down the center-line — but when the pilot went to rotate, he wouldn’t have been able to.
One pilot who experienced this CL600 glitch quickly noticed that his flight director was no longer in go-round mode but was matching the airplane’s level attitude. He was quick-witted enough to press the G/A switch and got off the ground. Another noticed the A/P was engaged before brakes release. Many CL600 pilots will have their right hand on the throttles and left hand on the nosewheel tiller until after 80 knots (with PNF holding in any aileron correction for crosswind). This leaves little time for PF to discover that his yoke’s frozen in pitch. Some experts say that the break-out clutch in the mechanical circuit between the Sperry A/P servo and the elevator cable drum should have saved this crew. There may be a ground-air sensing circuit in play here, however. That same ground-air sensing ensured that only the latter 10 of the 43 seconds of sought info was recorded on their flight data recorder (FDR). Later 604’s and RJs turn their FDRs ON when the right engine starter is engaged (604) or any strobe light is switched on (RJ). The accident FDR was not the same model that CL600s come from the factory with, says the NTSB. Time will tell if icing or automation switchology was the kill-joy here. Recall the ASW, Jan. 24, 2005, theory of what may have happened with CWS (control wheel steering) in the Flash 737 crash. See the AeroFlot crash article below. The autopilot can sometimes be an imbuggerance (i.e. quietly, impassively and modishly doing you more harm than good).
For details, visit: http://www.ntsb.gov/ntsb/GenPDF.asp?id=NYC04IA054&rpt=fa; for information on the Baotou CRJ200 crash, visit: http://www.iasa.com.au/CRJ200.htm