NTSB Factual Report

DAMAGE TO AIRPLANE

On October 28, 2007, a US Airways Boeing 757-200, N919UW, experienced a fracture of the truck beam on the left hand (LH) main landing gear (MLG) while preparing for departure from Charlotte/Douglas International Airport (CLT). The airplane was loaded with flight crew, cabin crew, passengers and cargo at the time of the truck beam failure. There was no other damage to the airplane and there were no injuries to any of the occupants or the ground personnel.

The fracture occurred forward of the pivot bore, and the truck beam was removed from the airplane and submitted for a metallurgical examination. The truck beam had accumulated 24,025 cycles since new and 2,707 cycles since its last overhaul which occurred in May, 2005.

TESTS AND RESEARCH

The ID surface displayed the presence of organic finishes consisting of an enamel topcoat and primer system, along with a layer of a corrosion inhibiting compound similar in appearance to Cosmoline. Quantities of grease were observed on the lower inside surface, in the vicinity of the pivot bore. The organic finishes within the 3:00 – 6:00 region (looking aft), exhibited deterioration in the form of blistering and lifting similar to that seen when exposed to paint stripping solutions.

Chemical analysis of the various compounds found within the truck confirmed the presence of both Royco 11MS (specified for pivot joint) and BMS 3-33 types of greases. The corrosion inhibiting compound was confirmed to be consistent with Cosmoline. Analysis also identified the presence of BMS 3-11 hydraulic fluid (Skydrol) in the grease and Cosmoline samples, as well as within samples of enamel and primer which had lifted and separated from the substrate.

Severe blistering and separation of the enamel topcoat was visible throughout the 3:00 – 6:00 region. This loss of enamel exposed the underlying primer which exhibited less advanced damage in the form of visible, widespread, small “bubbles” indicative of separation of the primer from the base metal.

The condition of the topcoat and primer was consistent with deterioration resulting from prolonged exposure to Skydrol which is known to act as a stripping agent for these finishes. The mixture of greases found in the truck beam would not have contributed to the finish degradation.

The fracture origin region was visible. Initiation occurred at the ID surface, near the approximate 6:00 position, and slow crack growth progressed through the thickness of the lower wall and in the circumferential direction to a final length of ~0.85”. Final fracture was due to ductile separation through the remaining truck beam wall.

Detailed examination of the origin region revealed a discolored/corroded thumbnail shaped region immediately adjacent to the ID surface that measured 0.05” long in the circumferential direction, and ~0.02” in depth. This region exhibited a topography consistent with propagation by a stress corrosion cracking mechanism which initiated at the ID surface of the truck beam. Beyond this region, the surface features were indicative of continued propagation by a transgranular fatigue mechanism to form a second thumbnail shaped area with a circumferential length of ~0.075” and depth of ~0.035”. Subsequent propagation switched back to stress corrosion cracking through the thickness of the wall and in a circumferential direction for a length of ~0.85”until final fracture occurred by ductile separation.

Scanning electron microscope examination of the fracture surface confirmed the propagation modes described above and that initiation also occurred by both fatigue and stress corrosion mechanisms from corrosion pits that were present on the ID surface. The pits were isolated features, surrounded by base metal surfaces which displayed clearly defined circumferential sanding or blending marks.

The enamel and primer were stripped from the ID surface in the fracture initiation area to expose the underlying cadmium plated base metal. Numerous pits were present under the protective finishes. In this area, the cadmium plating thickness measured 0.0001” to 0.0003” which is below the required minimum 0.0005”. The base metal after a complete stripping of the cadmium plating exposed additional scattered colonies of pits that had been obscured by the protective finishes and not visible until the enamel and primer were removed.

The general condition of the base metal was consistent with the surface texture and finish created by a sanding/blending operation. These features would typically not be visible on high strength steel parts because they are obliterated by the peening (either shot peen or roto-peen) which is the last required operation before cadmium plating.

The enamel and primer were stripped from the ID and OD surfaces at additional locations near and away from the fracture to expose the cadmium plating and base metal. At all locations examined, the thickness of the cadmium plating was below the minimum required 0.0005”, and the ID base metal exhibited the same sanded/polished appearance described above. Additional scattered colonies of corrosion pits were also present under the cadmium plating at random locations throughout the ID surface.

Metallographic cross sections were prepared through some typical pits which revealed their depths to be 0.001” – 0.004". Electron microprobe analysis indicated that cadmium and primer were present within the pits. These observations indicate that the pits were present when the last

application of protective finishes occurred.

Hole drilling measurements were completed to evaluate the residual stress state at the blended ID surface near the fracture origin. The average peak residual compressive stress measured ~140ksi and occurred at a depth of ~ 0.005”. At the origin location the average peak residual compressive stress measured ~68ksi and occurred at a depth of ~0.0035”.

For comparison, the surface was blended to remove 0.0025” of material, and the residual stress profile was reevaluated. The material removal resulted in a reduction of residual stress magnitude and depth, and a profile that was very similar to that at the fracture origin region.

The truck beam wall thickness at the fracture initiation area measured 0.29” which is within drawing tolerances. Chemical analysis and hardness testing (53Rc average) verified that the truck beam was fabricated from the required material in the specified heat treated

condition.

ADDITIONAL INFORMATION

Overhaul records for the subject part indicate that the inner diameter (ID) was blended in the area of the fracture to remove corrosion. The sanding marks observed on the ID surface traversed over the tops of, and through the corrosion pits, which indicates that the pitting was present at the time of blending. That blending operation, in the absence of subsequent peening (either shot peening or roto-peening), would create the observed ID surface conditions and would result in the reduced compressive residual stresses found in the fracture initiation area.

NTSB Probable Cause

The truck beam fracture initiated by fatigue and stress corrosion cracking from pits that were present on the ID surface of the truck beam, under the protective finishes. Crack propagation occurred by alternating modes of stress corrosion cracking and fatigue until final fracture occurred by ultimate ductile separation. The condition of the surface finishes and base metal at the fracture initiation area indicated that the pitting was present before the last application of cadmium plating and primer, rather than the result of corrosion since the last overhaul. The observed pitting appeared to be remnants of incomplete corrosion removal during the last overhaul.