Summary
On 27 April 2003 at 1319 hours UTC, a Robinson R22 Beta helicopter, registration G-LEDA, was involved in an accident near Royston, Hertfordshire. The helicopter was operating on a private flight with one pilot and one passenger.
Flight Details
The flight was planned to land in a client's back garden. As the aircraft approached the downwind leg at approximately 300 to 500 feet above ground level (AGL), the pilot heard a loud "grumbling" noise and felt a "twitch" to the left. Assuming an engine failure, the pilot lowered the collective and entered autorotation while attempting to increase speed to 75 knots. The pilot then turned into the wind and raised the collective to flare the aircraft. However, the right-hand skid of the helicopter touched the ground first, causing the helicopter to roll over onto its right-hand side.
Passenger Observations
The passenger noted that the clutch light had illuminated for a few seconds periodically during the flight.
Examination of the Helicopter
The wreckage was examined by the AAIB with assistance from the manufacturer. The helicopter had struck the ground with a high vertical speed, with the right skid slightly low, compressing both skids and the cabin seat structure before rolling onto its right side. Evidence of rotation on both rotor blades and the drive train at impact indicated that the engine had still been running.
One of the transmission V-belts was missing, and rubber deposits were found around the transmission compartment, suggesting the belt had flailed after failure. Small fragments of the missing belt were recovered. The remaining belt was intact and undamaged. No damage was observed on the sheave rims, and the actuator was found in a mid-position. No other pre-impact failures were evident.
Belt History
The helicopter had accumulated 2132 hours at the time of the accident. The V-belts had been replaced 12 hours prior, at 2120 hours. Belts are maintained on condition and have no finite service life. They are manufactured and fitted as matched pairs and are broken in for five hours at full RPM and tension before shipment.
Discussion of Belt Failures
Previous failures have not been associated with manufacturing or quality issues. The manufacturer identified that failures almost always occur with relatively new belts having less than 50 hours in service, most with less than 20 hours. Initiation is thought to be a belt strand coming out of a groove or rolling over in a groove, leading to an overload condition that tears the belt apart. Previous problems have been associated with combinations of high or above gross weight power applications, sheave alignment at installation or alignment shifts from initial belt wear-in, sheave surface condition (new belts on worn or corroded sheaves), actuator tension out of specification, or excessive belt slack at initial engagement. New belts are soft and tend to grip the sides of the grooves, increasing their tendency to climb out of the sheave.