HISTORY OF FLIGHT On September 8, 1994, at 1043 mountain daylight time, a Eurocopter AS 350 B2, N99TV, was destroyed when it impacted trees during descent at Cascade, Colorado. The pilot and one passenger received minor injuries, and another passenger was not injured. Visual meteorological conditions prevailed. The following is based on an interview with the pilot and the pilot/operator report. The helicopter landed at the 12,800 foot level of 14,110 foot Pikes Peak and two passengers, a reporter and photographer, disembarked to cover a news story. While awaiting their return, the pilot decided to reconnoiter another landing site in case it would be needed for a high altitude, high gross weight departure. During the takeoff, he made a power check and noted all instruments were normal. After determining it would be safe to depart from the original site, he returned and landed. After the passengers boarded, the pilot lifted the helicopter to a hover and made another power check, noting he had "plenty of power." With bleed air off, he brought the rotor speed up to 385 RPM (385 to 395 RPM normal), and noted Ng (free turbine rotation speed) was 99.7% (98% maximum continuous), torque was 86%, and T4 (exhaust gas temperature) was 720 degrees. He made a "left pedal turn" and initiated a climbout. When the helicopter was about 80 feet above the ground the helicopter yawed. He knew the engine was decelerating and confirmed this by noting a decay in rotor rpm, an Ng drop to 97.2%, and a drop in T4. The low rotor rpm warning horn also sounded. He decided his safest course of action would be to autorotate straight ahead. The helicopter collided with trees and rolled over on its right side. The engine was shut down by closing the fuel valve. TESTS AND RESEARCH Examination of the helicopter disclosed the Fuel Control Unit (FCU) at the flight gate position indicated 49 degrees, and the Power Turbine Governor Anticipator indicated 42 degrees. According to the Turbomeca Engine technical representative, the readings should have been 52 degrees and 50 to 60 degrees, respectively. The rigging measurements, he said, were similar to those of a "B" model helicopter, not a "B2" model. Fuel samples taken from the airframe fuel filter and fuel tank contained visible contaminants. These samples were analyzed by two independent Arlington, Texas, laboratories --- Atomus Laboratory and Spectro, Inc. Both laboratory reports disclosed the presence of filter material and particles of metal, glass, sand, and plastic. The metals were identified as being copper, iron, nickel, and chromium (see report attached). According to Turbomeca's report, such contamination could have resulted in the fuel filter becoming clogged and bypassed. The report said that the Fuel Control Unit filter, "which is only a protection filter, leaves particles up to 20 microns to enter the FCU and to disturb its operation." The source of these contaminants was not determined. No reports of fuel contamination were received from other pilots using the same fuel vendor. The engine and FCU were examined and tested at Turbomeca's facilities in Grand Prairie, Texas, under the supervision of a Federal Aviation Administration inspector. The engine operated within all manufacturer's specifications except for the maximum Ng test. Maximum engine speed attained was 52,088 RPM (allowable range is 53,531 RPM to 53,322 RPM). The FCU was unable to attain a target fuel flow of 236 liters per hour (equivalent to 53,359 RPM), but instead attained 215.5 liters per hour (equivalent to 52,173 RPM). After recalibrating the FCU to settings found at the accident site (Alpha = 49 degrees and Beta = 88 degrees), a fuel flow of 214.5 liters per hour (equivalent to 52,063 RPM) was attained. Disassembly of the FCU revealed a clogged amplifier piston filter (p/n 0164247350), a scored isochronous piston (p/n 0164242440), and a collapsed temperature compensator (p/n 9560142660). According to Turbomeca's report, the score marks are "evidence of an operation with polluted fuel."