NTSB Factual Report

On October 16, 2003, about 1530 Pacific daylight time, a Beech A36, N700TT, made an off-airport forced landing near Lancaster, California, following a loss of engine power while in cruise at 10,500 feet. The airplane had been modified by Supplemental Type Certificate with the installation of an Allison 250-B17-F/2 turboshaft engine. The pilot/owner was operating the airplane under the provisions of 14 CFR Part 91. The private pilot, the sole occupant, was not injured; the airplane sustained substantial damage. The personal cross-country flight departed Fullerton, California, about 1435, en route to Fresno, California. Visual meteorological conditions prevailed, and a visual flight rules (VFR) flight plan had been filed.

The National Transportation Safety Board investigator-in-charge (IIC) interviewed the pilot. The pilot said that he heard a bang, and noticed a loss of vacuum. The engine lost power, but the propeller continued to windmill. He declared an emergency, and attempted to glide to William J. Fox Field (WJF) in Lancaster. He was unable to make it to the runway. The airplane collided with power lines and fell to the ground about 4 miles west of the airport.

The Federal Aviation Administration (FAA) accident coordinator examined the wreckage on scene. Both wings separated from the airframe. The propeller and its gearbox separated from the front of the engine.

Initial Exam

The IIC and an investigator from Rolls-Royce Allison examined the airplane and engine at Aircraft Recovery Service, Littlerock, California. They drained 625 milliliters of a clear fluid that had a smell and color similar to jet aviation fuel from the airplane's fuel filter.

Examination of the three bladed propeller showed that all blades were in the feather position as indicated by the feather position reference mark. All blades exhibited damage to the leading edge with one blade curled backward. There was black arcing to the area surrounding the VHF antenna mounting on the top of the airplane.

The IIC observed that the engine sustained mechanical damage to the forward section. The forward flange of the compressor front support was crushed and the inlet guide vanes were buckled. The external oil sump tank buckled up and to the left. The compressor bleed valve manifold adapter buckled with the bleed valve displaced and contacting the Pc line from the compressor scroll. The lower engine mounting bracket was fractured, and the engine was displaced downward.

The propeller reduction gearbox housing fractured outboard from the mounting flange. All mounting bolts were in place and secured to the accessory gearbox. The propeller reduction gearbox rear housing split in one location forward to the forward housing splitline. There was some localized separation of the forward and rear housing. The propeller governor was intact and remained secured to the forward housing mounting pad. The sun gear remained engaged with the power take off (PTO) gearshaft, but had been displaced upward toward the 12 o'clock position. The BETA rod extended through the inner diameter of the sun gear, and bent downward toward the 6 o'clock position nearly 90 degrees and the rod end fractured. The forward gear teeth did not exhibit any damage. The balance of the propeller reduction gearbox remained with the propeller and attached to the propeller flange. The planetary gears, ring gear, spur gear, and balance of the propeller box components remained within the housing with no obvious visual damage observed.

Both the power and condition lever actuating cables fractured and separated at the fuel control and propeller governor end. The cables had been separated from the swaged inserts leading to the mounting brackets leading to the controls.

A manual check of all fuel, lube, and air lines for security determined that they were at least finger tight.

The coordinator assembly was intact. The actuating lever was bound in place, and indicated 0 degrees.

The N1 shafting system would not move when investigators attempted to rotate it via the first stage compressor wheel.

The N2 shafting system would not move when investigators attempted to rotate it via the fourth stage turbine wheel and the sun gear.

The IIC inspected the upper magnetic chip detector and noted that it was free of metal, but it did contain an amount of coke. There was slight varnishing of the anode and cathode.

The IIC inspected the lower magnetic chip detector and noted that it contained a substantial amount of metallic particles. The particles did bridge the gap from anode to cathode. A continuity check on the plug showed ground. Investigators then reconnected the chip detector to the wiring harness, grounded it against the airframe, and applied the airplane's power. The annunciator panel revealed several lights to include master caution and oil chips, which the Rolls-Royce representative said indicated that the chip light system was functional at the time of testing.

During engine removal, oil drained from the oil inlet line from the airplane's bulkhead to the accessory gearbox for approximately 2 minutes. Investigators drained the oil into a plastic container. They did not measure the oil quantity, but it appeared clean and did not smell burnt. They drained an additional 150-200 ml of oil from the lower chip port of the accessory gearbox. As with the oil inlet line, this oil appeared clean and did not smell burnt.

The IIC decided to ship the engine to Rolls-Royce Corporation, Indianapolis, Indiana, for further examination.

Indianapolis Examination

This engine examination occurred on January 11, 2004, under the supervision of the FAA.

Clean oil leaked from the exhaust collector after technicians placed the engine on a turnover stand.

The No. 8 oil supply line was crushed and broken. Technicians removed the oil pressure standpipe, and observed a moderate amount of coking inside the tube. However, they poured oil from a can through the tube and noted no obstructions.

The external oil sump tank was buckled and deformed. Oil was present in the tank and when rotated inverted, oil drained from the tank for approximately 10 seconds.

The beta rod was still engaged to the engine and extended through the inner diameter of the sun gear. The rod bent approximately 90 degrees at the point where the rod exited the sun gear.

The sun gear was still engaged to the power takeoff gearshaft and was displaced downward on an approximate 30 degree angle. Technicians noted static imprints (marks) that they felt were from contact with the inner diameter of the power takeoff gear. After removal of the sun gear, N2 remained bound in place.

After removal of the compressor, the N1 remained bound in place.

Removal of the horizontal fire shield showed a localized exit wound to the gas producer support in the plane of the first stage turbine wheel. The support was torn open from the 5:00 to 8:30 position. The gas producer support to power turbine support splitline separated in a localized area, and two of the splitline bolts fractured.

During removal of the turbine with the N1 and N2 couplings held in place (engaged to the compressor and pinion gear respectively), technicians noted that the internal retaining ring and flat washer from the No. 5 bearing were dislodged. The bearing remained in place, but was heavily damaged. The pinion oil nozzle that was mounted adjacent to the No. 5 bearing showed marks from impact with the retaining ring and/or washer. The power turbine rotating lab seal appeared to be worn. Technicians removed the bearing and associated components and placed them in a ziplock bag and transported to the Rolls-Royce Materials and Processes lab for further examination.

Technicians inspected both the N1 and N2 couplings. The N1 was unremarkable; the N2 was in good condition except for an approximate 120-degree section of the rear Teflon seal that was missing.

Technicians removed the No. 6/7 bearing external oil sump can; no oil was present in the sump and coked oil could be observed when visually inspected with a flashlight. The No. 6/7 bearing scavenge line also showed coked oil. The No. 6/7 bearing pressure oil line was free of obstructions.

The No. 8 bearing pressure oil strut was moderately coked on the inner diameter of the line, but the line flowed freely when flowing oil through the strut using an oil can.

Removal of the horizontal fireshield revealed a localized area of the gas producer support to power turbine support splitline separated from the 5:00 to 8:30 position. Two (2) splitline bolts fractured.

The No. 8 bearing sump nut showed damage from contact by the gas producer tiebolt; the tiebolt did not penetrate the wall. The snap ring was dislodged from the castellations, but had not completely released.

Removal of the thermocouple harness showed damage to the probe at the 6:00 position. The wire loop broke in two with a section missing and was not recovered. The base of the sheathing was also missing a small section. A second probe was slightly bent and misaligned but intact. The damage to the harness precluded functional (jet calibration) testing.

Technicians conducted a pressure check of the No. 8 sump. During this attempt, they noted heavy leakage at the 10:00 position.

The struts of the Gas Producer support at the 3- 9- & 12-o'clock positions cracked at the support cast I.D. radius.

Inspection of the No. 6 bearing showed the bearing fractured and contained a darkened appearance. The rollers were worn with the wear appearing heavier on one side. The No. 7 bearing contained a dry appearance and was difficult to rotate. Some coked oil was in the cavity. The No. 8 bearing was intact and functional. The Nos. 6, 7 & 8 bearings did not appear varnished or damaged as a result of inadequate oil supply. All bearing passages were clear, and the inside of the gearbox as well as the No. 1 and No. 2 bearings were oil wetted. The Nos. 3 and 4 bearings, located on either side of the pinion gear, were on position and intact but difficult to rotate. Visual inspection revealed metallic grit in the area of the respective separators.

Technicians opened the accessory gearbox; they removed and inspected the main oil delivery (Piccolo) tube. The primary passage was free of obstructions and the o-ring was intact.

Technicians removed the Gas Producer Tiebolt and found it in three pieces with fractures on both sides of the bore of the GP wheels. They noted that the tiebolt nut was loose with bowing of the retaining plate from rearward movement of the interstage seal.

The first stage turbine wheel (P/N 6886407, S/N X125872) contained an approximate 120-degree section of the wheel missing. The fracture appeared to originate near the rim of the wheel then propagate under the rim toward the web across the balance ring then turn tangential back toward the rim. The remaining airfoils were badly damaged with the majority of the airfoils missing.

The second stage turbine wheel showed rub from contact with the second stage nozzle diaphragm where the diaphragm moved aft coincident with failure and release of the first stage turbine wheel. The airfoils all sustained damage primarily to the leading edge. The 2nd stage turbine nozzle diaphragm and associated vanes were also heavily damaged.

The third stage turbine wheel contained heavy damage to the leading edge of the airfoils. Several airfoils showed material removal extending from the shroud tip. A large majority of the shroud was missing from heavy rub with the nozzle blade track.

At the conclusion of the engine disassembly examination, several component parts were boxed and transported to the Rolls-Royce Corporation Materials and Processes Laboratory for further investigation.

The engine logbook records contained no chip light history leading up to the time of the accident. On October 5, 1995, at 457.0 hours since new, an entry noted engine removal and shipment for oil contamination with repair actions listed under Dallas Airmotive work order no AG9199. On October 2, 1997, at 892.8 hours time since new, an entry noted replacement of the No. 1 Bearing and Carbon Seal by NAC work order NH6R616J. The maintenance records reviewed noted no additional work, other than routine inspections, on the engine.

Metallurgical Report

Rolls-Royce completed a metallurgical examination and submitted a report, which the Safety Board Materials Laboratory reviewed. Pertinent parts of the report follow.

First Stage Turbine Wheel

Three sections of the first stage turbine wheel separated. The stub shaft also fractured and separated, and a portion of the fractured tie bolt remained in the wheel. The metallurgist numbered the fractures from 1 to 4.

Fluorescent penetrant inspection (FPI) revealed multiple indications on the leading and trailing edge sides of the first stage turbine wheel. The metallurgist took blacklight images of the FPI indications. He observed eight indications on the trailing edge rim and five indications on the leading edge rim.

The metallurgist observed that fracture No. 1 was oxidized, but a small area near the trailing edge corner exhibited characteristics of fatigue. This area extended inward for approximately 0.07-inch followed by an interdentric fracture morphology. He examined the fracture with a scanning electron microscope (SEM). He observed oxidized, flat, transgranular crack propagation in two areas, which he said was consistent with fatigue. He observed characteristics of interdentric fracture in two areas.

Fractures Nos. 2, 3, and 4 exhibited fatigue. Closer examination of one of the fractures revealed that the largest area of fatigue was near the trailing edge of the wheel. This area extended around the rim face to a depth of 0.125-inch before transitioning to creep. The crack extended about 0.300-inch below the rim. SEM examination of the fracture surface revealed striations indicative of thermal fatigue. Close-up examination revealed a fracture morphology indicative of interdentric creep type fracture.

All of the first stage turbine wheel airfoils exhibited damage. The stub shaft fractured approximately 1.6 inches from the end. SEM examination of the fracture revealed characteristics indicative of tensile overload, and was considered secondary damage.

The metallurgist made a radial section near the mid-chord through a representative first stage turbine wheel airfoil. It revealed gamma prime solutioning to 0.180-inch outboard of the rim and extending to the airfoil tip. He also noted incipient melting near the fracture surface. This indicated exposure of the material inboard of the fracture to operating temperatures exceeding 2,100 degrees Farenheit (F). A rim crack exhibited transgranular propagation and alloy depletion along the crack consistent with thermal fatigue.

The hardness in the rim area and material type conformed to the engineering drawing.

The metallurgist felt that the first stage turbine wheel rim separated due to interdentric creep crack. He attributed this to elevated temperatures outside its normal operating temperatures consistent with a hot start.

Gas Producer Turbine Support Assembly and Energy Absorption Ring.

The gas producer support sustained damage to the outer wall and flange. The energy absorption ring was deformed. The first stage wheel created a hole in the side of the turbine support, but the separated pieces did not penetrate the energy absorption ring.

No. 8 Bearing and Rotating Labyrinth Seal

The labyrinth seal exhibited wear and impact damage. This was consistent with secondary damage occurring during the failure sequence. The No. 8 bearing rotated freely, and its components appeared in good condition. All bearing balls met the dimensional requirements.

First Stage Turbine Nozzle Assembly

The cooling holes for the first stage nozzle assembly exhibited mechanical damage, but no evidence of cracking. Radial sections made through three of the holes revealed no evidence of diffused phosphorous or phosphorous in the oxide on the surface. This indicated that there was no oil fire.

Second Stage Turbine Wheel

All of the airfoils on t

NTSB Probable Cause

A loss of engine power due to thermal fatigue failure of the first stage turbine wheel resulting in release of a section of its rim. The thermal fatigue was due to the engine exceeding its temperature limits during one or more start cycles.