WPR14TA236

The commercial pilot was conducting a nighttime law enforcement patrol flight in the helicopter. The pilot reported that, about 1.4 hours into the flight, the engine lost all power, and he subsequently performed an autorotation to an unlit empty field. The helicopter landed hard, and the main rotor blades subsequently struck and severed the tailboom. Postaccident examination of the engine revealed excessive wear in the turbine-to-compressor coupling components. The turbine-to-compressor coupling had moved, which caused its forward splines to contact the spur adapter gear shoulder and its rear splines to disengage from the turbine-splined adapter. The disengagement caused the decoupling of the turbine section from the compressor section and the subsequent loss of engine power. The excessive wear was determined to be the result of a misalignment of the engine's centerline shafting components. Two areas of misalignment were found. The first misalignment was caused by the incorrect placement of two shims at the interface between the rear diffuser and gearbox. A review of the engine maintenance logbooks revealed that maintenance personnel had placed these shims during the last compressor section overhaul, about 2 years or 1,432.1 flight hours before the accident. The second misalignment was found between the forward and rear pilot diameter bores of the exhaust collector assembly, which exhibited excessive runout when compared to engineering drawings. The maintenance records showed that, about 9 months before the accident, the exhaust collector assembly was reworked in accordance with the engine manufacturer's maintenance manual, which did not call for a runout check of the forward-to-rear pilot diameters but rather a runout check of both pilot diameters independently with reference to the bearing bore. At the time of rework, the runout-to-bearing bore values were within tolerance. Therefore, it is likely that the discrepancy was latent, predating the accident by a long period of time, and was not significant enough to cause previous shafting misalignment. Although neither of the misalignments individually were likely significant, cumulatively, they produced sufficient misalignment to cause the observed wear.

The helicopter was manufactured in 1995 and had been owned and operated by the City of Mesa through to the accident. It was equipped with a Rolls-Royce M250-C20R/2, 450 shaft-horsepower gas turbine engine, serial number CAE-295293. According to maintenance records, at the time of the accident the engine had accumulated 15,349.7 hours total time, and a Phase 4 (150 hour) inspection had been performed 8.5 hours prior. On June 8, 2014, at 0056 mountain standard time, a McDonnell Douglas Helicopters 369E, N504MP, landed hard following a loss of engine power during cruise flight in Mesa, Arizona. The helicopter was registered to and operated by the City of Mesa (Police Department) as a public aircraft flight. The commercial pilot and commercial pilot rated tactical flight officer were not injured. The helicopter sustained substantial damage to the tailboom during the accident sequence. The local flight departed Falcon Field Airport, Mesa, Arizona at 2315 the night prior. Night visual meteorological conditions prevailed, and no flight plan had been filed. The flight was a routine patrol, and about 1.4 hours after departure, the pilot performed a series of flight checks in the vicinity of Phoenix-Mesa Gateway Airport. Once complete, they flew to the north and a short time later while travelling 60-70 knots, and about 700 feet above ground level, the pilot heard an uncharacteristic change in engine tone accompanied by a decrease in rotor speed. The pilot began to diagnose the problem, and a series of caution warning lights illuminated followed by the engine-out tone. The pilot immediately initiated an autorotation, aiming for a tree-lined alfalfa field to the north. The helicopter cleared the trees, and prior to initiating the flare the pilot realized he needed to turn on the landing lights in order to properly see the ground. He was unable to turn them on in time, and he initiated the flare, landing the helicopter hard and level on the skids. The main rotor blades subsequently struck and severed the tailboom just aft of the tailboom mounting bolt access doors. The engine and airframe were examined by the NTSB investigator-in-charge, and technical representatives from Rolls-Royce, McDonnell Douglas Helicopters, and the Mesa Police Department. The engine remained firmly attached to the airframe, and did not display any indications of catastrophic damage. All fuel, oil, and pneumatic lines were intact and remained attached to their respective fittings, and the engine controls were continuous to their respective cockpit controls, and appeared to operate correctly. Fuel was present in the airframe fuel pump, and in the fuel line at the engine fuel nozzle. Engine Examination The engine was subsequently removed and examined by the group. The compressor section along with the diffuser was removed and the compressor rotor assembly could be rotated freely by hand; all observed bearings appeared intact, the rotor and impellor moved smoothly, and no obviousindications of damage were noted. The gearbox and accessory housing were removed, and both the power turbine and gas generator drive trains could be rotated smoothly, with corresponding movement of the accessory drive couplings and oil pump. Disassembly revealed no indication of catastrophic damage, and all gears and bearings appeared coated in oil. During removal of the gearbox, the five sets of shims at the interface flanges between the gearbox and rear diffuser were examined and compared to the shims documented in the engine logbook at the compressor section overhaul in April 2012 (1,432.1 flight hours prior to the accident). It was discovered that a 0.006 inch shim had been installed at the "C" (7 o'clock) position although the logbook stated a 0.004 inch shim was installed. Likewise, a 0.006 inch shim was installed at the "D" (11 o'clock) position, when the logbook stated a 0.008 inch shim was installed. The maintenance manual required that the installed shims have a minimum thickness of 0.002 inches, and a maximum total thickness of 0.02 inches per flange. According to Rolls-Royce, the shimming procedure is utilized to provide correct alignment of the compressor rotor to the gas producer turbine rotor, and thereby reduce the possibility of accelerated spline wear on the compressor to turbine shafting components. Examination of the turbine section revealed that the power turbine stage was locked, and could not be rotated by hand; although the gas generator stage could be rotated, it felt rough and produced a squealing sound. The turbine section was dissembled, and all bearings, wheels, and rotors were examined. No significant anomalies were noted, with the exception of the first stage wheel, which exhibited thermal discoloration and separation of trailing edges of all airfoils at the tips. With the major sections of the engine separated, the turbine-to-compressor coupling (centerline shafting) components were examined. Visual examination revealed that the mating spline teeth of the turbine splined adapter at the second stage wheel had worn away about 0.2 inches from its end, halfway across the spline contact face. Corresponding spline wear was present on the inner splined surface of the turbine-to-compressor coupling, about 0.25 inches inwards (halfway across its normal contact face.) The damage was consistent with forward displacement of the coupling, which exhibited dark discoloration from the spline-damaged end, and forward through 2/3 of its length. In an effort to explain the displacement of the turbine-to-compressor coupling, the spur adapter gearshaft (SAG) was removed and examined. The aft spline section of the SAG appeared intact, however the spline shoulder which would normally be circular, exhibited uniform 0.06 inch-deep spaces of localized wear on the aft side. The shape of the material loss matched the shape of the turbine-to-compressor coupling internal splines. Under normal alignment conditions, the turbine-to-compressorsplines would not come into contact with the SAG shoulder. Metallurgical and Metrological Examinations A group of engine components were sent to Rolls-Royce for metallurgical and dimensional analysis, the results of which were reviewed by specialists within the NTSB Offices of Research and Engineering, and Aviation Engineering. Analysis revealed that the microstructure, core hardness, case depth, and composition of the turbine splined adapter and the turbine-to-compressor coupling were as specified on the engineering drawings. Examination further confirmed that the shoulder on the SAG just forward of the aft splines had worn from contact with the turbine-to-compressor coupling internal splines. Additionally, the failed first stage turbine wheel blades exhibited thermal distress consistent with temperature exposures above 2,100 degrees F. Examination of the forward and rear pilot diameter bores of the exhaust collector assembly revealed that they exhibited a runout value about 80 percent beyond tolerance when compared to engineering data. Maintenance records indicated that the exhaust collector was last reworked in September 2013 in accordance with the Rolls-Royce maintenance manual. The manual did not call for a runout check of the forward to rear pilot diameters, but rather a runout check of both pilot diameters independently with reference to the bearing bore. In both instances, at the time of rework the runout to bearing bore values were within tolerance. An offset analysis was performed in an effort to evaluate whether the misalignment within the exhaust collector, along with the incorrectly placed shims could cumulatively create the wear mechanism observed in the turbine coupling and adapters. Results revealed that under these conditions, the hardened spline of the turbine-to-compressor coupling was working against the unhardened shoulder of the SAG shaft at levels about 25% higher than design criteria. Therefore, progressive wear similar to that observed on the SAG shoulder was likely. Prior Occurrences Rolls-Royce reported being aware of two prior events which exhibited similar decoupling of the turbine from the compressor due to misaligned shafting components. The first event was investigated by the NTSB (Accident Number CHI02LA163). The cause was attributed to improper maintenance of the exhaust collector, which appeared to have become deformed during an event prior to the accident, and subsequently exhibited multiple dimensional discrepancies to the forward pilot bores. The second event did not result in an accident, so was not investigated by the NTSB. Rolls-Royce reported that this occurrence was the result of an incorrectly reworked exhaust collector. They were unable to locate specific information as to how it came to be out of limits, nor where they provided with hardware conditions, maintenance actions, or overhaul activities that might have contributed to the misaligned condition. Rolls-Royce reported that the event likelypre-dated the manufacture and installation of the accident engine.