"We have nothing at the impact sites to say this is definitively what it is." Maybe the NTSB should look again - over and through the Denver dilemma. It may not be as nitty gritty and transparent a solution as flying grit.

A Convenient but non-Credible Sleuthing?

In Air Safety Week dated 26 February ("Thirteen plus One") we reported upon some of the likely causes of the 22 windshields cracking on 14 aircraft during a 90 minute period at Denver International Airport on 16th February last. The Denver-based NTSB lead investigator Jennifer Kaiser has now come up with an FOD explanation that she thinks "flies". It's a natural phenomenon according to Jennifer. "The only commonality across aircraft type, operator, location, time and phase of flight was the wind and weather," Kaiser claimed. Despite aircraft being parked in different orientations, fine particles of grit were being swept along by gusts as high (at one stage) as 48 mph claims Kyle Fredin of the National Weather Service. On one parked aircraft, a CRJ 700, all four cockpit windows were cracked. Either the grit, or the wind, was turning corners.

SkyWest Airlines appeared to suffer the most cracked windscreens, with nine of its CRJs and Embraer Brasilias affected. Windshield fractures developed on six planes as they were taking off, some of which aborted; on one just after it had landed; on two as they were taxiing to the terminal after landing; on three as they were parked at the gate; on one as it was being pushed back from the gate; and on one while it was at 19,000 feet.

The main imponderable was that the crazed windscreens appeared on a number of different types of aircraft. The Denver Post Report however doesn't go into any finer detail except to say that the fine particles caused a fine pitting that in turn must have caused the cracking. There's no common denominator disclosed such as whether or not all the windscreens involved had thermostatic feedback mechanisms made by company X. You will recall that the ASW surmise was that a cold-soaked thermostat may not have been able to kick-start its feedback process and thus allow the heating current's inflow to be modulated. No feedback on increasing temperature of the NESA warming layer equates to a system determination of "obviously needs more heating current", which is supplied and the result is a quickly (and non-uniformly) heated NESA layer - which ultimately causes the outer layer to craze. In part, as ASW surmised, this can also be caused by the vice-like grip of the windscreen's cold-soaked frames and the many bolts clamp-holding at least three layers of transparencies that are being non-uniformly heated. Allowing 30 minutes for the cabin's APU heated cabin air to bring up the windscreen's cold-soaked metal frames and bolts to a uniform temperature with the glass is seen as the first step of a logical process of pre-heating.

Speaking as one who has flown in hail, freezing rain, snow and desert sandstorms in jets with heated windscreens, at speeds over 480mph, as a convenient solution for the 14 jet happenstance at DIA, it's all a bit see-through. For pressure of particles on a windscreen, think proportional to the square of the indicated airspeed, not just ten times Denver's peak gusts of 48 mph. AD 2007-04-09 affecting the EMB135/145 (and wef 23 Mar 07) addresses poor electrical contact due to loosening of the attaching hardware of the power cables of certain windshield temperature controllers.... to prevent overheating of the power cable terminals of the windshield temperature controllers.

The cracking, said Kaiser, occurred on the outer of the three layers of the windshields; microscopic analysis showed fine particles caused pitting that in turn caused cracking. Investigators were unable to determine the precise nature of the debris as there were no "transfer" marks of the material onto the windshields (think CSI), Kaiser said. "We have nothing at the impact sites to say this is definitively what it is." Maybe the NTSB should look again - over and through the Denver dilemma. It may not be as nitty gritty and transparent a solution as flying grit.