HISTORY OF FLIGHT On September 3, 2004, about 1630 Pacific daylight time, a Beech F33A, N1854N, operated by the pilot-owner, experienced a pitch control system malfunction approaching the Mariposa-Yosemite Airport, Mariposa, California. The pilot diverted to the Fresno Yosemite International Airport, Fresno, California. On landing flare to runway 29L, the problem manifested itself again when the pilot attempted to perform a normal landing flare. The pitch control movement was inhibited. The airplane impacted the runway hard in a nose low attitude and was substantially damaged. Neither the private pilot nor passenger sustained injuries. Visual meteorological conditions prevailed. The personal flight was performed under the provisions of 14 CFR Part 91, and it originated from Santa Ana, California, about 1400. The pilot reported to the National Transportation Safety Board investigator that he first detected a problem approaching the Mariposa airport. When the landing gear was extended and the wing flaps were partially extended, his ability to move the control yoke was noticeably restricted. The pilot reported that he tried to pull back on the yoke and use the manual trim to regain control, but the airplane did not respond. Upon experiencing this event he diverted to Fresno, which has a larger airport. Approaching Fresno, the pilot alerted the air traffic controllers about his problem, and they notified the respective crash, fire, and rescue facility. According to the pilot, he lost pitch control when he attempted to flare during the landing. The airplane abruptly nosed down and impacted the runway surface, collapsing the nose gear and bending the fuselage. A small ground fire started beneath the airplane, but it was immediately extinguished. TESTS AND RESEARCH Under the direction of the National Transportation Safety Board investigator, the airplane was examined by Federal Aviation Administration (FAA) personnel at the accident site and again following its recovery. In summary, at the accident site the control yoke's movement was found restricted. Only partial up-elevator travel could be obtained. With the exception of components associated with the autopilot, no anomalies were found with the airplane's impact damaged structure or systems. The FAA reported that an examination of the autopilot pitch control servo revealed that one side of the servo bridle cable had broken and was detached from the servo capstan bridle cable locking pin. The other servo bridle cable was found attached to the main elevator control cable and servo, but it was twisted and deformed. Both servo capstan bridle cable guide pins were bent full forward. One nylon guide pin guard was missing and was found in the fuselage belly. The Edo Air Mitchell Pitch Servo (actuator) model #1C750-1-668P, serial number 543, was removed from the airplane and examined under FAA supervision at a repair station. According to the FAA, the servomotor initially would not start running until approximately 12 volts of direct current (DC) were applied to the motor. The servo is designed for 28-volt operation. After several on-and-off cycles of DC, the motor started running with approximately 5 to 6 volts DC applied. However, even at the full 28 volts of DC, the motor operated slowly. The servo gear train actuator mechanism was found sticky. Repair station personnel reported that the servo needed maintenance. Airplane Information. The airplane's current owner reported that his airplane was maintained on an annual inspection basis. The last inspection was accomplished on October 1, 2003, about 11 months before the accident. Since that date, the airplane had been operated about 57 hours. The airplane had been manufactured in 1981 by Beech, and the autopilot was installed as original equipment. FAA personnel reviewed the airplane's maintenance records. No autopilot maintenance records were located. Autopilot Information, Century Flight Systems, Inc. Management personnel at Century Flight Systems reported that the model of autopilot installed in the accident airplane had not been manufactured for several years. Upon request, Century Flight Systems provides instructions to industry for its continued airworthiness (ICA). Prior to the September 3, 2004, accident, the ICA appropriate for use in maintaining the accident model of autopilot had not specifically addressed (1) the time interval for the autopilot's recurring inspections; (2) examination for proper cable winding on the capstan; (3) examination for proper travel of controls from limit to limit; and (4) examination for fraying of the bridle cable. On February 16, 2005, Century Flight Systems issued revision "G" to its ICA, and on February 21, 2005, Century issued service bulletin #CSB-2005-01. The addresses in this bulletin were "All Century Dealers, Service Centers and Aircraft Owners of Century Autopilots." In the bulletin the company indicated that "... a situation occurred involving a bridle cable that became frayed and broke at the capstan pin causing some control wheel friction. Upon investigation it was determined that the bridle cable had not been wrapped correctly and continued excessive unwrapping of the bridle cable from the capstan had caused it to fray and eventually break." Accordingly, Century Flight Systems recommended that "all bridle cables should be checked on each annual inspection and/or every 100 hours of operation." In addition, Century Flight Systems recommended that a physical examination be performed of all bridle cables that had not been recently checked for fraying, and thereafter, that the aforementioned recurring inspection schedule be adhered to. Also recommended in the service bulletin was a check of the bridle cable wrap on the capstan and travel when the control wheel is moved from mechanical limit to limit to ensure that there is still cable wrap on the capstan at each limit.