Mitsubishi MU-2 Marquise

The flight departed VFR, and when near the destination airport, was advised by air traffic control to hold VFR over the "plaza" and to make left 360 degree orbits. Several witnesses reported light rain was occurring at the time of the accident; there was no lightning or thunder. One witness located where the airplane came to rest reported observing the airplane emerge from the base of the clouds in a 45-degree left wing low and 20 degrees nose low attitude. He momentarily lost sight of the airplane but then noted it rolled to a wings level attitude. He also reported hearing the engine(s) "cutting in an out." Another witness located approximately 1/4 mile north of the accident site observed the airplane flying eastbound beneath the clouds in a right wing and nose low attitude, he also reported hearing the engine(s) sounding like they were "cutting in and out." A pilot-rated witness located an estimated 1,000 feet from where the airplane came to rest estimated that the ceiling was at 300 feet and there was light drizzle. He observed the airplane in a 45-degree angle of bank to the right and in a slight nose low attitude. He stated that the airplane continued in that attitude before he lost sight of the airplane at 250 feet. The airplane impacted trees then a concrete wall while in a nose and right wing low attitude. The airplane then traveled through automobile hoists/lifts which were covered by corrugated metal, and came to rest adjacent to a building of an automobile facility. Impact and a post crash fire destroyed the airplane, along with a building and several vehicles parked at the facility. Examination of the airplane revealed the flaps were symmetrically retracted and landing gears were retracted. No evidence of preimpact failure or malfunction was noted to the flight controls. Examination of the engines revealed no evidence of preimpact failure or malfunction; impact and fire damage precluded testing of several engine accessories from both engines. Examination of the propellers revealed no evidence of preimpact failure or malfunction. Parallel slash marks were noted in several of the corrugated metal panels that covered the hoists/lifts, the slashes were noted 25 and 21 inches between them. According to the airplane manufacturer, the 25 inch distance between the propeller slashes corresponds to an airspeed of 123 knots. Additionally, the power-off stall speed at the airplanes calculated weight with the flaps retracted and 48 degree angle of bank was calculated to be 122 knots. Review of NTSB plotted radar data revealed that the pilot performed one 360-degree orbit to the left with varying angles of left bank and while flying initially at 1,300 feet, climbing to near 1,500 feet, then descending to approximately 800 feet. The airplane continued in the left turn and between 1502:10 and 1502:27, the calibrated airspeed decreased from 160 to 100 knots. At 1502:27, the bank angle was 48 degrees, and the angle of attack was 26 degrees. Between 1502:30 and 1502:35, the true heading changed indicating a bank to the right. The last plotted altitude was 200 feet, which occurred at 1502:35. A NTSB weather study indicated that at the area and altitude the airplane was operating, NWS VIP level 1 to 2 echoes (light to moderate intensity) were noted. Additionally, the terminal aerodrome forecast (TAF) for the destination airport indicated that temporarily between 1400 and 1800 (the flight departed at approximately 1436 and the accident occurred at approximately 1503), visibility 5 miles with moderate rain showers, scattered clouds at 1,500 feet, and a broken ceiling at 3,000 feet.

The airplane was on final approach to land at Hilton Head Airport, when according to witnesses, it suddenly rolled to the right, and descended, initially impacting trees at about the 70-foot level, and then impacting the ground. A fire then ensued upon ground impact, and the debris field spanned about 370 feet along an azimuth of about 082 degrees. Examination of the airplane wreckage revealed that left wing flap actuator and jack nut measurements were consistent with the wing flaps being extended to 40 degrees, and on the right wing the flap jack nut and actuator measurements were consistent with the right flap being extended to about a 20-degrees. In addition, the right flap torque tube assembly between the flap motor and the flap stop assembly had disconnected, and the flap torque tube assembly's female coupler which attaches to the male spline end of the flap motor and flap stop assembly was found with a cotter pin installed through the female coupler of the flap stop assembly. The cotter pin, had not been placed through the spline and the coupler consistent with normal installation as per Mitsubishi's maintenance manual, or as specified in Airworthiness Directive 88-23-01. Instead, the cotter pin had missed the male spline on the flap motor. In addition, the flap coupler on the opposite side of the flap motor was found to also found to not have a cotter pin installed. Company maintenance records showed that on April 3, 2001, about 87 flight hours before the accident, the airplane was inspected per Airworthiness Directive (AD) 88-23-01, which required the disassembly, inspection, and reassembly of the flap torque tube joints. In addition, on July 9, 2001, the airplane was given a phase 1 inspection, and Bankair records showed that a company authorized maintenance person performed the applicable maintenance items, and certified the airplane for return to service.

The pilot was maneuvering the airplane south of the airport preparing to make a VFR approach. Witnesses observed the airplane in a right spin. NTAP data showed the airplane to be well above stall speed before disappearing from radar. Examination of the radar data revealed that in 6 seconds, ground speed dropped 31 knots, from 200 knots to 169 knots, and altitude dropped 440 feet, from 11,760 feet to 11,320 feet (4,400 feet per minute). In the next 6 seconds, ground speed dropped another 31 knots, from 169 knots to 138 knots, and altitude dropped 1,020 feet, from 11,320 feet to 10,300 feet (10,200 feet per minute). According to the manufacturer, if the throttles were to be brought back into Beta (flat pitch) range, it is possible that one propeller could go into Beta an instant before the other propeller. If this were to happen, the airplane would instantly snap roll and enter a spiral. The pilot had received an estimated 4 hours of dual instruction in the airplane.

Visual meteorological conditions prevailed for the planned cross-country flight for which the pilot obtained a weather briefing, filed an IFR flight plan, and received an ATC clearance. Approximately 8 minutes after takeoff, radar indicated the airplane was at 11,200 feet msl, heading 241 degrees, with a ground speed of 180 knots. No distress calls or additional communications with the pilot were recorded, and radar contact was lost. The airplane impacted the ground in an uncontrolled descent. The right wing tip tank separated from the airplane and was found 0.18 nautical miles from the main wreckage. The teardown and examination of both engines disclosed that the type and degree of damage was indicative of engine power section rotation and operation at the time of impact. There were no complete systems intact at the accident site due to the impact sequence and post-impact fire which consumed the aircraft. The landing gear and flaps were found in the retracted position. The portion of the right propeller shaft coupling found at the site was fractured through 360 degrees. Metallurgical examination revealed that the propeller shaft coupling failed in fatigue. The presence of the fatigue cracks indicated the coupler fractured in fatigue in service, and the fatigue cracks were not the result of ground impact. The circumferential fracture intersected the ends of several internal spline teeth. The origin of the fatigue crack could not be determined because of severe corrosion damage on the fracture surface. Fatigue propagation was in the aft direction and from the inside to the outside of the coupling. The engine core rotating components would have bee free to rotate when uncoupled from the propeller shaft. The maintenance records indicated that the failed coupling had accumulated approximately 4,000 hours since new, and 1,250 hours since engine overhaul in 1989. Since 1990, as a result of fatigue fractures, the manufacturer introduced several design changes for the propeller shaft coupling via optional Service Bulletins to be accomplished at the next access or hot section inspection (HSI). Impact and thermal damage of the right propeller precluded a determination of the in-flight blade angles. The calculations by the airplane manufacturer indicated that "the [intact] airplane was capable of continued flight" with the right propeller feathered, and that the "airplane can keep attitude, but cannot climb and cannot maintain altitude" with the right propeller in the flat pitch or wind milling positions, respectively. Metallurgical examination of the component brackets and associated bolts from the right tip tank revealed the separation of the tip tank resulted from a single-event overstress fracture of both the forward and aft tank attachment fittings. Calculations showed that a 3.763 radians per second (35.9 RPM) spin rate would cause the failure of the forward wing fuel tank attachment fitting. There had not been a previous in-flight separation of a wing tip fuel tank on this model airplane.

While on final approach to Caucasia Airport, the twin engine aircraft crashed in unknown circumstances 3 km short of runway. The aircraft was destroyed and the pilot, sole on board, was killed.