On July 10, 2001, about 0630 eastern daylight time, a Bell 206B helicopter, N3298G, was substantially damaged during an autorotation after a loss of tail rotor thrust in Secaucus, New Jersey. The certificated commercial pilot was not injured. Visual meteorological conditions prevailed at the time of the accident. No flight plan had been filed for the local flight, which began in Jersey City, New Jersey. The aerial application flight was conducted under 14 CFR Part 137. According to the pilot, he had just completed a spraying pass, and was heading westbound about 100 feet and 60 mph, parallel to power lines, and 15-20 feet above trees. He suddenly felt something "break", and the helicopter started to yaw. The pilot realized that the helicopter had experienced a tail rotor failure, but he didn't want to autorotate into the trees. He reduced the throttle slightly, maintained altitude, spun around two to three times, and cleared the trees. He then headed for a gravel area and rolled off the throttle, but couldn't stop the spin. He attempted to cushion the landing, but the helicopter hit the ground hard. After landing, the pilot pulled the circuit breakers and turned off the master switch. He then noticed fuel on the ground and exited the helicopter. Power to the tail rotor was provided through a series of driveshafts, connected by "Thomas couplings." According to the Federal Aviation Administration (FAA) inspector who examined the wreckage, the Thomas coupling disc assembly on the aft end of the aft short shaft was twisted and bent, and the tail rotor drive shaft just aft of the coupling was bent and separated, about mid-length. In addition, an AN174-7A retaining bolt was missing from the coupling. The missing bolt and washers were found in the helicopter's tailrotor drive shaft tunnel area, and appeared to be undamaged. The self locking retaining nut for the bolt was not located. Pertinent tail rotor drivetrain components were forwarded to the Safety Board's Materials Laboratory. These included the aft short shaft, which was still connected to the first segment of the tail rotor drive shaft by one bolt through a Thomas coupling, and the "missing" bolt. According to the metallurgist’s factual report: "A Thomas coupling...of the type involved in this examination consists of a series of thin plate rings (discs) with bolt holes at 90-degree intervals. Bolts through one set of diametrically opposing holes are attached to a flange on a forward shaft, and bolts through the other set of opposing holes are attached to a flange on an aft shaft. The tail rotor drive system has nine of these couplings. The discs transfer torque while allowing for some amount of angular displacement. The Thomas coupling at the forward end of the aft short shaft was intact and did not contain any visible cracks. The individual discs of the coupling were slightly gapped apart between bolt locations, and no damage was noted on the splined adapter on the forward side of the Thomas coupling. The aft short shaft was intact, but the shaft portion contained a slight overall bend. ...The Thomas coupling at the aft end of the aft short shaft...was received without one of the bolt and nut sets that attach the coupling to the splined adapter at the forward end of the first segment of the tail rotor drive shaft. The individual discs of the coupling were gapped apart between bolt locations, and the outer discs contained visible wrinkles. No evidence of cracking was noted in the discs. The Thomas coupling [at the forward end of the first segment of the tail rotor drive shaft] has been rotated 180 degrees (around the bolt opposite the missing bolt) from its assembled position. The forward face of the [flange] adapter contained a series of rotational scoring and impact marks from contact with the bolt heads and discs. The pattern of this damage was as if the short shaft was rotating around the bolt opposite from the missing bolt. Additional examination of the Thomas coupling and associated components revealed that (1) portions of the normal seating positions for the head of the missing bolt and the washer were relatively free of dirt and grease, and (2) the bore surface of the hole for the missing bolt in the flanged adapter at the forward end of the first segment contained thread imprint marks. The nuts for the coupling bolts are lock nuts and are specified to have a minimum of 3.5 inch-pounds of running (tare) torque. Procedures reportedly specify application of 50 to 70 inch-pounds of torque above the tare torque when the bolt and nuts are installed. The torque needed to initiate removal of the nuts from the three remaining coupling bolts was measured. Loose dirt was cleaned from the protruding threads before torque measurements. The following...shows the torque measurements from the coupling nuts. Bolt and nut 2 are from the position opposite from the missing bolt. Bolts and nuts 3 and 4 are from the other two positions. The scale on the torque wrench used was divided into increments of 5 in-lbs. Therefore, measurements at values in this range are not precise." The removal torque and tare torque from bolt/nut 2 were "about 5 in-lbs." The removal torque from bolt 3 was 55 in-lbs, and the tare torque was "less than 5 in-lbs." The removal torque from bolt/nut 4 was 50 in-lbs, and the tare torque was "10-20 in-lbs." "The nut from bolt 2 was assembled onto the recovered bolt (assumed to be bolt 1, the missing bolt from the Thomas coupling.) Measurement of the tare torque of nut 2 on bolt 1 gave a value of 5 to 10 inch pounds. Three pieces of the No. 3 bearing hangar were received for examination.... Approximately one fourth of the hangar was missing. Examination of the fractures in the hangar pieces revealed features typical of overstress separations. No evidence of fatigue or other type of progressive cracking was noted. Both segments of the tail rotor drive shaft contained substantial compression buckling on a spiral plane adjacent to the fracture area. Fracture features and associated deformation on the shaft segments were typical of fracture as a result of application of excessive torsional and bending loads. The direction of torsion was as if the forward portions of the segments rotated clockwise (aft looking forward) relative to the aft portions of the segments." According to maintenance records, the tail assembly was changed 57 flight hours before the accident. The maintenance technician reported that he reused the bolts and self-locking nuts. According to FAA Advisory Circular 43.13-1, the reuse of the nuts was permissible, provided the nut threads were in good condition.