NTSB Factual Report

On November 17, 2003, at 2105 eastern standard time, a Hughes TH-55A, N9240D, piloted by a private pilot, was substantially damaged during an emergency landing after takeoff from a private helipad located about 3.8 nautical miles northwest of Purdue University Airport (LAF), Lafayette, Indiana. The pilot reported a loss of directional control while climbing through 400 feet above ground level (AGL) at 45 knots. Night visual meteorological conditions prevailed at the time of the accident. The 14 CFR Part 91 personal flight was not operating on a flight plan. The pilot received serious injuries. The flight was originating at the time of the accident and was en route to LAF.

The pilot stated that he departed for LAF to perform night take-off and landings for "night qualifications." About 400 feet AGL and 45 knots, "it felt as if someone had grabbed the tail and pulled it to the right, then pushed it to the left." He then heard a "pop" as though the cabin was struck with an open hand. This sound was followed by sound comparable to "numerically controlled milling machines... Instantly, I was in a very rapid rotation to the right." Everything was black except a few lights along 475 west, which flashed like strobe lights with each rotation. He stated that the lighted panel helped him focus on the rotor rpms. There were no warning lights.

He stated that he dropped collective and rolled the throttle into the override position in order to stop spinning. He stated that letting the throttle off of the override immediately sent him into a right spin. He then rolled the throttle back into the override position, which immediately arrested the rotation. During the descent, he let the engine out of override and pulled full collective after the landing light began to illuminate the ground,

The 1988 Hughes TH-55A, serial number 67-15414, helicopter was purchased from the military through the Defense Reutilization and Marketing Service on January 22, 1996. According to the Federal Aviation Administration (FAA) inspector, there were at least two civilian owners since the helicopter's conversion to civilian use.

Inspection of the helicopter by a FAA inspector revealed the tail boom right cluster fitting was fractured. A piece of the right tail boom attachment fitting and two pieces of the right rear cluster fitting clevis were sent to the National Transportation Safety Board (NTSB) Materials Laboratory for examination.

According to the NTSB Materials Laboratory Factual Report, which is available in the docket of this report, the tail boom attachment fitting was bent in the inboard direction of the helicopter. The fracture face displayed a rough grainy surface, features consistent with a bending overload event, and an angular change of the surface, consistent with a shear lip, along the upper inboard edge.

The pieces of the right rear cluster fitting were portions of the lugs from the tail boom support strut clevis. The fracture face on the upper lug of the clevis revealed a rough textured surface consistent with an overload event and no indications of any deformation at the edges. Much of the perimeter of the fracture face was darker than the interior of the fracture. Scanning electron microscope examination of the fracture face revealed the surface features consisted of a mixture of ductile dimples and cleavage. (Ductile dimples are an indication of overstress fracture and are cuplike depressions created as a result of microvoid coalescence; cleavage is a brittle fracture mode that progresses on preferred planes within individual crystals in the microstructure). Measurements revealed the upper piece of the clevis had a thickness of 0.080 inches and is identified as 'thin lug' or 'BSC' fittings that range in thickness between 0.070 - 0.110 inches. Later versions or the cluster fittings are identified as 'thick lug' fittings and range in thickness between 0.115 - 0.155 inches.

The lower lug consisted of a flat lug portion, similar to the upper piece, with a small portion of the vertical web, located on the inboard side of the cluster fitting. The fracture progressed through both the lug portion and the vertical web portion. The lug portion was severely bent in the upward direction adjacent to the lug fracture. In this area, two small portions were partially separated from the parent piece and also bent upwards. Examination of the fracture face revealed a darker perimeter region similar to what was observed on the upper lug.

The vertical web had been weld repaired in the fracture area, and fracture features were found only in two zones that covered less than half of the cross section of the vertical web piece. The remainder of the web surface had a smooth weld surface that did not contain any fracture features. The upper fracture zone displayed blue coloration and the lower displayed brown coloration. A sphere of weld material was found on the fracture surface between the two fracture zones. Hardness measurements were taken and found to have an average Hardness, Rockwell "B" scale of 57 (equivalent to a tensile strength of 291,000 psi).

Examination of the lug portion of the lower clevis lug revealed evidence of a welding arc strike on the underside that displayed an annular shape with a depressed porous center that could be produced by a welding rod using the shielded metal arc welding (SMAW) process.

According to the manufacturer, the cluster fitting, fitting fuse, and frame joint is specified in design as a 4130 steel casting. The fitting is to be heat treated. The fitting is to be heat treated to HRB 88 to Rockwell "C" scale of 27 (equivalent to a tensile strength of 85,000 - 126,000 psi). Specified weld repairs are allowed on the tube sockets only.

According to the FAA inspector, a review of the helicopter's maintenance logs revealed no entries pertaining to welding of the cluster fitting.

NTSB Probable Cause

The fractured cluster fitting resulting in a loss of directional control during initial climb and the improper repair of cluster fitting by unknown personnel. The night conditions were a contributing factor.