HISTORY OF EVENT Use your browsers 'back' function to return to synopsisReturn to Query Page
On May 31, 2002, approximately 0755 Pacific daylight time, a McDonnell Douglas MD-11F, N608FE, registered to Wilmington Trust Company, operated by Federal Express Corp., and crewed by two airline transport rated pilots, incurred minor damage during a windshield fire while standing prior to pushback at the north cargo ramp of Seattle-Tacoma (SEATAC) International Airport, SeaTac, Washington. Both pilots were uninjured. Visual meteorological conditions existed and an IFR flight plan had been filed for the impending flight. The intended flight, which was a regularly scheduled, domestic cargo flight, was to be operated under 14 CFR 121, and the aircraft was destined for Oakland, California, following its planned departure from Seattle. The aircraft had arrived from Memphis, Tennessee, earlier in the day.
The crew reported that prior to pushback, maintenance personnel were engaged in the repair of a small (1 x 3 inch) hole on the underside of the left horizontal stabilizer. The Captain reported that following his preflight he executed the standard cockpit "flow" checks during which he determined that the windshield Anti-Ice was OFF and the windshield De-Fog was ON. Although the engines were not running the aircraft was powered up electrically. During this time the flight crew observed the co-pilot's forward windshield panel (R1) crack and observed what they described as flames emanating from the lower portion of the windshield. The captain perceived that the flames were on the exterior of the windshield and the first officer opened his right window (R2) and used a hand held halon fire extinguisher to attempt to extinguish the fire. The captain radioed SEATAC ground control advising them of the fire condition. The aircraft was then de-powered and the fire went out.
The R1 windshield panel was removed and replaced along with its associated anti-ice controller. The windshield was examined by the investigator in charge and then shipped to the manufacturer for further examination (refer to photographs 1 and 2). The anti-ice controller was also examined and then shipped to the Federal Express avionics department and then to the manufacturer for additional examination.
The R1 windshield panel was reported to have been manufactured on July 28, 1998. The R1 windshield heat controller (part number: SYLZ-51737, serial number: 64F) was manufactured in November of 1991. Its total time since new was reported as 22,756 hours and the unit had undergone 5,306 cycles (refer to Attachment SA-I).
TESTS AND RESEARCH
The R1 (copilot's forward) windshield panel, serial number 98287H0515 was removed and shipped to the manufacturer (PPG Industries, Inc.) for a more detailed examination and evaluation. A visual examination revealed several delamination islands (voids) along the upper edge of the windshield. The delaminations were irregular in shape and lacked any coloration and, they occurred at the interface of the outer ply inner surface (anti-ice heating film layer).
The fracture was observed to originate at the lower edge of the windshield just forward of the aft right - lower corner and was characteristic of a thermally related fracture. There was no evidence of moisture ingress or other discrepancies in the vicinity of the fracture origin. Additionally, there was no evidence of fire (refer to Attachment PPG-I).
The windshield heat controller was shipped to the facilities of Smiths Aerospace for testing and examination. The controller was examined and tested on July 30, 2002. The unit failed the initial bench test and when an overheat condition was simulated it was found that the unit did not automatically trip off line. The unit was evaluated and found to have numerous failed components (refer to Attachment SA-I).
In a telephone conversation between the Investigator in Charge and the Smiths Aerospace representative present at the windshield heat controller unit testing the following information was provided:
1. When electrical power was initially applied to the controller in the "OFF" mode it was observed that the controller commanded the windshield anti-ice system to begin heating (uncontrolled outflow of 336 volts).
2. The circuitry within the controller had failed thus preventing the auto-heat control system from functioning and regulating the heating to the anti-ice panel.
3. The unregulated output commanded the anti-ice panel to increase in temperature ultimately to failure of the panel.
4. A short within the controller prevented the system from being de-powered as long as power was available to the aircraft systems (contacts C2 and C3 shorted, Q1 shorted and Q2 open), and
5. When aircraft power was removed (crew shutdown of external/APU power) power was then terminated to the windshield controller, which could no longer command full heat to the windshield.
6. The manufacturer reported that they had never seen this combination of failures within the heat controller during its previous history.