HISTORY OF FLIGHT On February 28, 1994 at 2305 Alaska standard time, a Lockheed L-1011 Tristar airplane, N303EA, operating as Rich Air Flight 303, was departing Fairbanks International Airport, Fairbanks, Alaska, and experienced a mechanical power loss on the number 3 and number 1 engines and an internal fire on the number 1 engine. The takeoff was aborted and during the taxi back, the fire on the number 1 engine was extinguished by the Fairbanks Fire Department. The positioning flight, operating under 14 CFR Part 91, was departing Fairbanks and the destination was Miami, Florida. Visual meteorological conditions prevailed and an instrument flight rules flight plan was filed. The accident occurred during the hours of darkness. The airplane received substantial damage and the crew of 14, the only occupants, received no injuries. According to the flight crew, they were beginning to advance their power levers from the stabilized power setting of 1.1 EPR when they heard a loud bang and saw a glow on the right side of the airplane. The right engine fire warning system activated and the crew energized the fire bottle for the number 3 engine and the fire was extinguished. While they were completing the aborted takeoff and engine shutdown they saw a glow on the left side of the airplane and then the number 1 engine fire warning system activated. The crew discharged a fire bottle for the number 1 engine and did not receive a secondary indication. As the airplane cleared the runway the fire department extinguished a fire in the tail cone of the number 1 engine. TEST AND RESEARCH Examination of the number 3 engine showed that the 6th & 7th stage, intermediate pressure compressor failed and had exited the engine case and cowling. Examination of the number 1 engine showed an entrance puncture on the inboard side of the cowling and that a piece of the 6th/7th stage wheel had penetrated the number 1 engine core. The airport was searched and all the pieces of the 6th & 7th stage compressor wheel were found except for a section that contained 2 1/2 holes of the 6th stage. All pieces were submitted for metallurgical examination. The 6th & 7th stage I.P. compressor rotor shaft is a rearward cantilevered assembly that is bolted to the 5th stage assembly through 25 holes in the rim of the 6th stage disk. The metallurgist's report shows that these holes had corrosion pitting and fatigue cracking. The deepest crack measured 0.13 inches deep and 0.23 inches wide centered in the bore of hole number 13, which is marked in the metallurgists report. According to Rolls Royce the crack would have to be at least 0.25 inches deep for a failure to occur at 100% N2 speeds. Their calculations also show that with a crack of only 0.13 inches deep, the N2 speed would have to reach 118%. According to the NTSB Laboratory, the N2 speed could not accurately be calculated because this engine has a 3 spool compressor, and the digital flight data recorder records only exhaust pressure ratio (EPR) and not an N2 speed. Examination of the flight engineer's panel showed that the N2 gauge for the number 3 engine had a tell tale marker which recorded an overspeed of 106%. The red line was 102.05%. According to the Captain and First Officer, their maximum EPR setting was 1.535 for all three engines. They had targeted to use 1.475 and the highest they saw was 1.3 to 1.35. According to the flight data recorder readout information, the highest EPR reached was on the number 2 engine with a maximum of 1.3758 occurring 13 seconds after longitudinal acceleration started. According to Rolls Royce Engines, they had previously identified the cracking in the holes of the 6-7 stage I.P. compressor disc. Two Service Bulletins (SB) were released which called for the severe reduction in the life limit of the rotor shaft assemblies and discs. The regulatory authorities in the United Kingdom, the location where the engines were manufactured, have made compliance with the S mandatory two years earlier. According to the Federal Aviation Administration the operators do not have to comply with a Service Bulletin. Rich International Air had not complied with these Service Bulletins. ADDITIONAL INFORMATION Supplements C, D, and I were not completed because there was no airplane wreckage and the cockpit was secured. The N2 gauge was removed and calibrated, however, the shop which calibrated the gauge failed to indicate if the gauge was reading high or low. According to Rolls Royce, in their comments about observations and metallurgical examinations, without the portions of disc containing the last 2 1/2 bolt holes, it is difficult to say if the failure was caused by overspeed or fatigue. Rolls Royce believes that all the facts to date point to an overspeed condition.