2018 Pilatus PC-6 accident at Grenoble Isère (F-BTCG)

Accident overview and flight sequence On 16 March 2018, a Pilatus PC-6/B2H2 registered F-BTCG departed Grenoble Isère aerodrome for a post-maintenance test flight. The pilot was accompanied by an aircraft mechanic. After reaching a clear area south of the aerodrome, the pilot began the planned test maneuvers. Following a stall maneuver, the pilot found that rudder pedals had no effect, while aileron and elevator control remained available. The pilot informed the aerodrome controller and returned to land. The aircraft entered a long final for paved runway 09. Because the nature of the damage was unknown, the pilot chose to land with flaps retracted. Alignment with the runway centerline proved difficult. At about 300 ft, the pilot decided to land on the adjacent unpaved runway 09R. During a very short final, at the flare and power reduction at roughly one to two meters above the ground, the aircraft rolled right then left; the wingtips and propeller struck the ground. The aircraft cartwheeled, exited the unpaved runway to the left, and came to rest on the right edge of the paved runway. ## Aircraft condition and pilot experience The aircraft was heavily damaged. Wingtips, rigging wires, and propeller blade tips were damaged. The rear fuselage near the fin was bent and partially destroyed, and the tailwheel was partly torn off. The pilot was 62 years old, held a PPL(A) and a PC-6 type rating, and had about 2,000 total flight hours, including about 500 hours on the PC-6, mainly for parachute drops. ## Meteorological conditions At the time of the accident, conditions at the aerodrome were: wind 070° at 3 kt, CAVOK, temperature 10 °C, and QNH 1003 hPa. ## Pilot testimony The pilot reported no anomalies during the preflight inspection. During the stall, no abnormal behavior was perceived, and the yaw-axis failure was detected only during turbine power-up. During final, the pilot had difficulty maintaining alignment and had to constantly correct the path. The pilot did not recall the final approach speed. At the flare, the aircraft began to “oscillate in roll” before a hard ground contact. The pilot intended to make a three-point landing as usual. ## Aircraft history and detailed examination The Pilatus PC-6 F-BTCG, serial number 551, was built in 1963. After a 1983 accident, it was rebuilt using parts from another Pilatus PC-6 (serial number 660, built in 1968), with reconstruction spanning 1984 to 1990. A general overhaul was completed in April 2013. At the time of the accident, the aircraft had 12,260 flight hours since new and 1,270 hours since the last general overhaul. A detailed examination with the manufacturer showed that the four rivets joining the lower collar supporting the rudder bellcrank to the rudder axis had sheared. The assembly did not conform to the design drawings: the lower collar was secured by only four rivets instead of eight. The upper collar assembly was also non-conforming, with only six rivets instead of eight and a mark corresponding to a pre-drilled hole. Laboratory tests indicated play between the collar and the rudder control axis, sudden shear failure of the rivets due to overload, and larger-than-expected diameter differences between holes and rivets, which over time could lead to play under vibration. ## Manufacturer information and maintenance The original design drawings for this assembly date from 7 October 1958, when the lower and upper collars were attached to the rudder axis using two conical dowels. In December 1960, the design was changed to eight rivets for both assemblies, requiring four additional holes. Further design changes occurred between 1962 and 1995. Strength calculations showed that eight rivets were necessary to provide an adequate safety factor against aerodynamic loads on the rudder; with only four rivets, shear failure and loss of rudder control could occur depending on configuration and flight phase. Maintenance before the accident did not involve the rudder axis and was unrelated to the in-flight failure. Records from 2003 to the accident did not mention replacement of this part; earlier records were unavailable. In 2014, during a general overhaul by another maintenance facility, replacement of a ball bearing on the rudder axis required removal and reinstallation of the upper collar and new rivets. The work order did not detail the operations, and it could not be determined whether the two additional holes (six instead of four) and the pre-drilling were done then or earlier. The aircraft logbook showed no events since the 1990 reconstruction that would have required replacement of the rudder axis. ## Findings and conclusion The rudder axis could not have come from either PC-6 serial number 551 or 660, both built after the design change to eight-rivet assemblies. The investigation could not determine the origin of the axis, but it is probable that the assembly was made during the 1984–1990 reconstruction period. Following the accident, the manufacturer issued Service Bulletin No. 27-006 in July 2017 for all Pilatus PC-6 aircraft, requiring inspection of the rudder axis and verification of assembly conformity, with instructions for non-conformities. The rudder control failed in flight, significantly reducing yaw controllability. This greatly increased the pilot’s workload and stress, making it difficult to maintain runway alignment during landing. Attention to alignment came at the expense of speed monitoring. It is very probable that the oscillations during the flare resulted from a low-speed stall.