On October 29, 1994, at 0745 hours Pacific daylight time, a Fairchild SA-227AC, N191AF, experienced a loss of aileron control shortly after departing runway 27L at Oakland International Airport, Oakland, California. The airplane was operating as Ameriflight (AMF) flight 1894, a nonscheduled cargo flight, under Title 14 CFR Part 135. The pilot had filed and began a company visual flight rules flight plan to Reno, Nevada. The airplane, operated by Ameriflight, Burbank, California, did not sustain any damage. The certificated airline transport pilot, the sole occupant, was not injured. Visual meteorological conditions prevailed. The pilot returned to Oakland International Airport without further incident. Ameriflight maintenance personnel told National Transportation Safety Board investigators that the right aileron cable broke at its 90-degree pulley attach points. They also said that the cable had accrued 15 hours since it was installed. The cable replacement was required by a Federal Aviation Administration (FAA) airworthiness directive. Safety Board investigators sent the broken aileron cable, with an exemplar from the cable stock used by Ameriflight maintenance personnel, to the Safety Board Metallurgical Laboratory. The Safety Board metallurgist reported that during examination of the cable, he found substantial bending deformation in a uniform direction in most wires next to the separation. He also said that a magnified optical examination revealed no evidence of wear or corrosion near the separation or anywhere else on the cable. The metallurgist conducted a scanning electron microscope (SEM) examination of the broken cable. The examination showed that some separated ends displayed necking down or fractures in a 45-degree plane. This condition is typical of overstress separations. Additionally, the metallurgist said, in part, that he . . .found many of the cable ends contained relatively flat fracture areas on a transverse plane, which could result from fatigue cracking. These flat surface areas usually terminated at a longitudinal fracture area . . . . A detailed SEM examination of the flat fracture area on these cable ends revealed ". . .a somewhat roughened and micro fissured fracture morphology, but no classical fatigue striations were found. Ductile dimples, a characteristic of overstress separations, were found in portions of the wire separations outside of the flat fracture areas. . . ." The cable material quality and strength conformed to the manufacturer's specifications. The operator reported that the mechanics and inspectors who changed the cable said the new cable could not have been kinked due to the procedure used. The mechanics said that the new cable end was attached to the old cable after cutting off the swaged-on fitting at the underfloor bellcrank end. The new cable is then pulled in by pulling out the old cable from the control yoke end. If the new cable were kinked, it would not pass through the keepers around the pulleys in the control circuit and other confined areas in the cable run. The operator did not say that during the final connection of the cable the attached ends could be twisted. The airplane manufacturer's representative told Safety Board investigators in a telephone interview conducted on December 7, 1994, that a kink could be induced during the installation process. He said that the maintenance procedures required the mechanic to swage the ends of the cable and conduct a pull-test greater than 1,200 pounds. He also said that if the cable was kinked, it would fail during the pull-test preinstallation.