Metal Fatigue Blamed for Helicopter Crash [Archive]

Lu Zuckerman

29th Jan 2004, 05:37

To: Time Out

Here is the technical information. The full report can be found here:

http://www.ntsb.gov/ntsb/brief2.asp?ev_id=20030616X00878&ntsbno=DEN03FA099&akey=1

The tail rotor parts were examined in Washington, DC, on June 18, 2003, at the National Transportation Safety Board's Materials Laboratory. The initial examination of the two trunnion pieces showed four separate fatigue planes originating from the inner splines.

Each fracture surface on the trunnion pieces had portions that were flat with curving boundaries and curving crack arrest lines, features typical of fatigue. Multiple fatigue origins and cracks were observed in the fracture surfaces of the two trunnion pieces, and at the roots of splines 9, 10, and 21. Fatigue fracture features emanated from multiple origins at the root of the contact face of spline 22. Fatigue cracks were also observed in splines 27 and 1.

During the detailed examination, trunnion piece number 1 was sectioned through the splines in a plane parallel to the outboard surface to obtain a polished cross-sectional profile of the splines. Surface roughness was examined at several spline roots. The average distance from peak to valley was approximately 4.53 microns. Dividing this value by 4 produces an estimate of the root mean square (RMS) roughness of 1.13 microns (44 microinches). According to the engineering drawing, the maximum RMS roughness is specified as 63 microinches.

Fourteen spline root radii were measured in the cross-sectional profile. The average measured radius was 151 microns (0.00595 inch). The smallest measured radius was 136 microns (0.00535 inch). These values were lower than the approximate value specified in the drawing (approximately 180 microns). However, according to the manufacturer, these values are within the range of values typically observed on trunnion assemblies, including those used for fatigue testing.

The trunnion piece was then cut perpendicular to the longitudinal axis in the bushing area, and core hardness and conductivity were measured. The average hardness was 67.1 HRB, and the conductivity was 39.6 percent IACS. According to engineering drawings, the trunnion is specified as aluminum alloy 2024-T351 with a minimum hardness of 68 HRB. According to the Aerospace Structural Metals Handbook (1995 Edition, CINDAS/USAF CRDA, Handbooks Operation, Purdue University), the typical conductivity for this alloy and temper is 29.7 percent IACS.

In the trunnion assembly, Pro-seal 890, Class B-2, is applied at the outer end. The sealant is specified to MIL-S-8802. Sealants made to MIL-S-8802 are specified to operate up to 121 degrees Celsius (250 Degrees Fahrenheit). The cover at the outer end of the accident trunnion assembly was removed from piece number 1. The sealant was brown, cracked, and degraded, consistent with exposure to temperatures above the specified range.

The manufacturer reported that following the initial examination, they conducted tests to measure hardness and conductivity changes in alloy 2024-T351 samples with a baseline hardness of 76 HRB and a conductivity of 29 percent IACS after exposure to elevated temperatures. After 300 hours at 200 degrees Celsius (392 degrees Fahrenheit), the hardness was measured as 74 HRB, and the conductivity was 39.4 percent IACS. In another sample, after 8 hours at 250 degrees Celsius (482 degrees Fahrenheit), the hardness was 67 HRB, and the conductivity was 41 percent IACS, values similar to those observed on the accident trunnion.

The manufacturer also reported that room temperature and elevated temperature fatigue tests were performed on alloy 2024-T351 samples heat treated for 8 hours at 250 degrees Celsius. The fatigue tests were conducted using notched coupons (Kt=3) in tension (R=0.1). The results showed that at room temperature, the fatigue endurance limit did not decrease for the heat treated samples relative to the untreated samples. At 200 degrees Celsius, the fatigue endurance limit was 23 percent lower than that of the untreated sample at room temperature. At 250 degrees Celsius, the fatigue endurance limit was 41 percent lower than that of the untreated sample at room temperature. However, the manufacturer stated that a 41 percent reduction in the fatigue endurance limit is insufficient to produce fatigue within the design load spectrum.

The tail rotor hub assembly, blade grip assemblies, and the tail rotor blades were examined. As assembled on the helicopter, the hub and blade grip bushings have an interior Ampep X1 liner that is located within an exterior corrosion-resistant steel ring. The inner surface of the Ampep X1 liner is polytetrafluoroethylene (PTFE) and glass fiber composite, and the remainder of the liner is a glass fiber reinforced composite. The total thickness of the liner is 0.3 millimeters (0.008 inch).

The hub bushings on the flap axis of the hub were examined. An area of sliding contact was observed on the inboard side of the hub bushing outer ends corresponding to contact with the trunnion assembly covers. The interior surfaces of the bushings appeared black in color around approximately 270 degrees. In an approximate 90 degree area near the blade axis, the interior surfaces of the bushings appeared reddish-brown intermixed with black dimples. Within approximately 0.2 inch of the outer ends of the bushings, the surfaces had a similar reddish-brown appearance all around the circumference. On the outboard side at the inner end of the bushing, the interior liner was worn away, exposing the exterior steel ring.

A portion of the outboard side of one bushing was sectioned and examined using a scanning electron microscope (SEM). Viewed using backscattered electrons, the surface in the black region showed broken glass fibers. The fibers were analyzed using energy dispersive x-ray spectroscopy (EDS), and the spectra showed peaks of calcium, silicon, aluminum, oxygen, and carbon, consistent with glass fibers from the exterior portion of the Ampep X1 liner.

The tail rotor blades are held to the tail rotor hub by tension-torsion straps. Motion between the blades and the hub occurs as the pitch of the blade changes. The inside diameter of the hub end of the blade grip contain two bushings, an "inner " bushing closer to the hub, and an "outer" bushing further away from the hub, that rub against the outer surface of the hub and transmit bending and torsional resistance loads. The blade B grip assembly was cut circumferentially to facilitate the examination of the outer blade grip bushing on that assembly.

The Ampep X1 liner of the inner bushing for blade B was worn through to the exterior steel ring at the hub edge of the bushing, inboard side (toward the tail boom). The liner for the outer bushing from blade B was also worn through to the exterior steel ring at the blade edge of the bushing, outboard side (away from the tail boom). Strips of white semitransparent thin film with the appearance of PTFE were observed at the steel bushings for the blade B grip assembly, particularly at the blade edge of the outer bushing. Similar film material was observed at the hub edge of the inner bushing for blade A.

National Transportation Safety Board Materials Laboratory Factual Report Number 03-108, October 28, 2003 is provided as an addendum to this report.

The tail rotor assembly was inspected at the 2,400-hour inspection on April 3, 2003. The assembly had a total time of 1,776.0 hours. The tail rotor trunnion was also inspected at that time. The trunnion had 547.8 total hours at the inspection. At the accident, the trunnion had 698.0 total hours. The published service life limit of the trunnion is 2,700 hours.

ADDITIONAL INFORMATION

Parties to the investigation were the FAA Flight Standards District Office, Salt Lake City, Utah, IHC Life Flight, Agusta Aerospace Corporation, and Turbomecca Engine Corporation.

The helicopter wreckage was returned and released to the operator's insurance company on November 5, 2003.

:E :E