DFW07IA184

Fatigue failure of the power turbine outer shaft for undetermined reasons.

Shortly after departing in a single-engine helicopter from a platform in the Gulf of Mexico, the commercial pilot heard a "high pitched winding noise" and a "pop" followed by a loss of engine power . The pilot landed the helicopter safely on the water with the floats fully inflated. Shortly thereafter, a large wave impacted the helicopter and broke out the right windshield and rolled the helicopter inverted. The pilot and the passenger were able to exit the helicopter and deploy one of the on-board life rafts. They were later rescued by a local boat and the USCG. Examination of the engine revealed the power turbine (PT) outer shaft was intact; however, there were two spiral fractures observed on the shaft. One spiral fracture was approximately 3-inches-long and originated between a crack within one curvic tooth and the mid-shaft. At the mid-shaft location, as the shaft wall thickens, the spiral direction changed 90 degrees and continued for approximately another half-inch. The second spiral fracture was approximately 180 degrees from the first and was approximately one half-inch in length with one end starting at the root of a curvic tooth. The loaded faces of the curvic-coupling teeth were galled. The fourth stage turbine connector nut was in its proper location and appeared to be undamaged. The deformable lock feature was still engaged with the nut. The easy removal of the nut from the shaft was hampered by the presence of a piece of curvic tooth fragment that was wedged between the nut and the shaft. Once the tooth fragment was removed, the nut rotated freely and was removed. The ring described by the forward face of the shaft, exhibited areas of polishing consistent with contact against the ring described by the aft face of the pinion gear. According to the Rolls-Royce metallurgical examination report, the fracture surfaces exhibited evidence of fatigue. The manufacturer also stated that there have been 16 other instances where the power turbine outer shaft had cracked since 1990; however, none of these resulted in an engine shut down. The only damage noted in these cases was fretting on the mating wheel. Each of the cracks initiated in the curvic coupling, and fretting was found at the fracture origin. The fretting was attributed to a loss of clamp load either from inadequate torque on the PT nut, or raised metal on the curvics that prevented proper seating. As a result of these 16 events, and prior to the PHI accident, Rolls Royce implemented several mitigating actions to prevent a future occurrence. These included changes to the assembly process, enhanced instructions in the overhaul manual on better handling practices to better protect the curvic couplings. The source of the fractures could not be determined.

HISTORY OF FLIGHT On August 16, 2007, at 1513 central daylight time, a single-engine Bell 407 helicopter, N433PH, was not damaged during an emergency landing to the water in the Gulf of Mexico. The commercial pilot sustained minor injuries and the passenger was not injured. The helicopter was registered to and operated by Petroleum Helicopters Incorporated (PHI), of Lafayette, Louisiana. A company visual flight rules flight plan was filed for the on-demand air taxi flight that departed the East Cameron 109 (EC109) offshore platform at about 1507, and was destined for the West Cameron 98 (WC98) offshore platform. Visual meteorological conditions prevailed for the flight conducted under 14 Code of Federal Regulations Part 135. In a written statement, the pilot reported that he departed the EC 109 platform and climbed to an altitude of approximately 1,000 feet and filed his flight plan with the operator's communications center. Approximately five minutes after departure, the engine chip light illuminated and the pilot immediately turned back to the departure platform. While in the turn, the pilot reported hearing a "high pitched grinding noise" and a "pop" before the engine stopped producing power. The pilot entered an autorotation to the ocean as he made a Mayday call to his company's communications center. When the helicopter was approximately 75 to 100-feet-high above the water, the pilot initiated a flare to slow the helicopter's descent and attempted to deploy the emergency floats; however, he did not pull the handle hard enough. The pilot made a second attempt, which was successful, and the floats deployed when the helicopter was approximately 10-feet-high above the water. The pilot "pulled pitch" and the helicopter landed safely on the water with the floats fully inflated. Shortly thereafter, a large wave impacted the helicopter and broke out the right windshield and rolled the helicopter inverted. The pilot and the passenger were able to exit the helicopter and deploy one of the on-board life rafts. Once both individuals were in the life raft, the pilot activated the EPIRB and another PHI helicopter was able to direct a nearby shrimp boat to the life raft. The pilot and the passenger were taken onto the boat and were later recovered by a United States Coast Guard helicopter. PERSONNEL INFORMATION The pilot held a commercial pilot certificate for rotorcraft-helicopter with a helicopter instrument rating. In addition, he was also a certificated flight instructor for rotorcraft-helicopter and instrument helicopter. The pilot held a current second-class FAA medical certificate and reported having accumulated a total of 1,855-hours, of which 139-hours were in the same make and model. METEOROLOGICAL INFORMATION The pilot reported the weather as wind between 180 and 220 degrees at 20, gusting to 25 knots, visibility 10 miles, broken ceiling 2,000 feet, and seas 6 to 10 feet. ENGINE EXAMINATION The engine was a Rolls-Royce M250-C47B, serial number S/N 847233. The M250 engine was a 2-spool turbo-shaft engine. Two power turbine (PT) stages drive the output shaft through a reduction gearbox. The gas producer consists of a compressor section, which has a single centrifugal stage driven by a 2-stage turbine. The combustion chamber is a cylindrical type. The engine power management system consists of a Hydro-Mechanical Unit (HMU), which is controlled by an Electronic Control Unit (ECU). According to PHI maintenance records, the engine had accumulated 8,606.49 hours time since new (TSN) and of the turbine assembly had accumulated 1,904.7 hours time since new, 282.03 hours time since overhaul (TSO), and 439 cycles since overhaul at the time of the accident. The engine was removed from the airplane at PHI facilities in Lafayette, Louisiana. The engine was partially disassembled under the supervision of the FAA with representatives from PHI, Bell, and Rolls Royce in attendance. Examination of the engine revealed signs of an uncontained engine failure near the area of the third stage turbine wheel. After the initial examination at PHI, the engine was shipped to Rolls Royce in Indianapolis, Indiana, where a Safety Board Power plants Engineer supervised a more detailed examination. According to the NTSB Power Plants Group Chairman report, the examination included the removal of the upper and lower gearbox chip detectors, which both contained metallic debris. This debris was also found in the N2 speed pickup, oilcan and the oil filter. Scoring was noted on the oil scavenge pump's #6 and #7 bearing scavenge gear elements and cavities. The #1 and #8 bearing scavenge gear elements and cavities were not damaged. The turbine module, which included the combustor and the gas producer turbine, were also examined. The combustor liner was intact. The fuel nozzle had previously been removed. The fuel nozzle swirler boss was cracked at the combustor chamber interface. The first stage nozzle shield was intact. The #8 bearing sump cover was deformed near the center consistent with contact with the gas producer turbine tie bolt. The third stage rotating labyrinth seal was intact. Approximately 150 degrees of the inner diameter was severely rubbed, which was consistent with contact against the #6 and #7 bearing housing. A 90-degree segment was worn completely through. The knife-edges were circumferentially and unevenly deformed, with an approximately 90-degree segment almost completely rolled-over. The PT inner shaft was intact; however, the inner race of the #6 bearing was severely deformed and spalled. The mid-shaft pilot ridge was deformed through approximately 120 degrees. The inner surface was corroded. The mating thread for the PT inner shaft spanner nut appeared to be undamaged. The PT inner shaft spanner nut appeared to be undamaged. The inner diameter of the PT inner shaft at the plane of the #6 bearing displayed rotational scoring as well as metal transfer, which was consistent with contact against the outer diameter of the turbine-to-compressor coupling. The PT outer shaft was intact; however, there were two spiral fractures observed on the shaft. One spiral fracture was approximately 3-inches-long and originated between a crack within one curvic tooth and the mid-shaft. At the mid-shaft location, as the shaft wall thickens, the spiral direction changed 90 degrees and continued for approximately another half-inch. The second spiral fracture was approximately 180 degrees from the first and was approximately one half-inch in length with one end starting at the root of a curvic tooth. The loaded faces of the curvic-coupling teeth were galled. The fourth stage turbine connector nut was in its proper location and appeared to be undamaged. The deformable lock feature was still engaged with the nut. The easy removal of the nut from the shaft was hampered by the presence of a piece of curvic tooth fragment that was wedged between the nut and the shaft. Once the tooth fragment was removed, the nut rotated freely and was removed. The ring described by the forward face of the shaft, exhibited areas of polishing consistent with contact against the ring described by the aft face of the pinion gear. According to the Rolls-Royce metallurgical examination report, the fracture surfaces exhibited evidence of fatigue. The gas producer tie bolt was fractured in two locations, which corresponded to the axial front and aft face planes of the first stage turbine disk segment. The outer case of the gas producer turbine bearing support was fractured in three locations. All three fractures were in the axial orientation and through the forward flange. One fracture had progressed through approximately 75 percent of the housing length and the other two had progressed through approximately 95 percent of the housing length. The #8 bearing cavity was wetted with oil and the bearing retainer sleeve appeared to be undamaged. All