On May 3, 2006, about 1900 eastern daylight time, a Hughes 269B, N9471F, piloted by a commercial pilot, was substantially damaged during a forced landing following a failure of the tail rotor drive shaft during cruise flight near Sullivan, Ohio. The personal flight was being conducted under 14 CFR Part 91 without a flight plan. Visual meteorological conditions prevailed at the time. The pilot was not injured. The local flight departed the pilot's private property near Sullivan, Ohio, about 1820. The pilot reported that the helicopter was in level flight approximately 600 feet above ground level (agl), when he heard a "very loud" bang. He stated that the aircraft began to yaw and he subsequently determined the aircraft had no tail rotor authority. He set up for a run-on landing to a field; however, he subsequently heard a "metal grinding sound" and a "high pitched squealing noise." He initiated an autorotation at that time. The pilot reported a "sensation of the tail boom swinging left [and] right, up [and] down." The helicopter encountered a ditch during touchdown, which collapsed the right skid. Upon exiting the helicopter, the pilot noticed that the left main tail boom support (cluster) fitting had broken loose from the frame. The tail rotor drive shaft and cluster fitting were examined by the National Transportation Safety Board Materials Laboratory. Damage to the tail rotor drive shaft was consistent with rotational contact between the drive shaft and another component. The drive shaft failed at the point where it passed through a clearance hole in the forward tail boom closure fitting. The fitting also exhibited features consistent with contact by a rotating component. The tail rotor drive shaft fracture surfaces were consistent with an overstress failure. The cluster fitting exhibited fracture markings indicative of fatigue. Fracture features on both lugs indicated localized fatigue origin areas on the outer surfaces of each lug. On the lower lug, the fatigue propagated nearly through the entire cross section before final fracture. On the upper lug, the fatigue penetrated approximately 35 percent of the cross section before final fracture. Surface corrosion was visible on the lower lug near the fatigue origin area. No corrosion existed at the fatigue origin on the upper lug. Federal Aviation Administration (FAA) Airworthiness Directive (AD) 2003-13-15 R1, which became effective August 10, 2004, was applicable to the accident aircraft. The AD noted that compliance was required in order "to prevent failure of a tailboom support strut or lug on a cluster fitting." The AD required modification or replacement of the original cluster fittings, part numbers 269A2234 and 269A2235, within 6 months or 150 hours time-in-service (TIS). The AD required dye penetrant inspections of the lugs within 10 hours TIS, and thereafter at intervals not to exceed 50 hours TIS, until the lugs were modified or replaced. Nominal measured lug thickness was 0.076 inch. This was consistent with cluster fitting part numbers 269A2234 and 269A2235. The failed lug did not appear to have been modified. Review of the aircraft logbook revealed that a 100 hour / annual inspection was completed on May 23, 2005, at 3,686 hours total aircraft flight time. The entry noted that AD 2003-13-15 R1 was complied with at that time. The airworthiness compliance record for that date noted AD 2003-13-15 R1, and stated that it had been complied with by performing a dye penetrant inspection without finding any cracks. The entry also noted that the modification kit was not installed at that time.