On January 21, 1997, about 0620 Pacific standard time, N75ZV, a Beech 1900D, operated by Mesa Airlines, Inc., as United Express Flight 7830, had a flight control malfunction during approach to the Seattle-Tacoma International Airport in Seattle, Washington. The airplane received minor damage and landed safely. The captain, first officer, and five passengers were not injured. Visual meteorological conditions prevailed and an instrument flight rules flight plan had been filed. The flight departed from Walla Walla, Washington, and was destined for Seattle. The scheduled domestic commuter flight was conducted under 14 CFR 135. In an interview with the Safety Board (record of interview attached), the captain stated that he began the day of the incident in Walla Walla, Washington, by performing a preflight inspection "by the checklist." The inspection included operating the flaps and running up the engines. The captain stated that "everything worked normally" during the inspection and run-up. The flightcrew then boarded the passengers and departed uneventfully from Walla Walla. The captain stated that he and the first officer briefed the instrument landing system (ILS) approach to Seattle, but planned on getting a visual clearance to the runway. Dark night conditions prevailed. The airplane broke out of the clouds about 4,200 feet, and the flightcrew was cleared for a visual approach to runway 16R. The captain stated that the first officer then called for "flaps 17." No problems were noted during the flap extension. The flightcrew then requested and received permission to land on 16L. The airplane was slowed to 140 knots, and the flightcrew was cleared to land on runway 16L about 3-1/2 miles from the airport. After getting lined up with the runway, the first officer called for "flaps 35." The captain lowered the flaps. He immediately heard a "whump" and saw the first officer move the control yoke. The first officer told him that he had a problem moving the controls. The captain stated that he did not want to have a transfer of control at that moment, so he elected to have the first officer continue to fly the airplane. The captain placed his hands on the yoke in an attempt to feel the control forces. He stated that the "controls felt restrictive." He elected to continue the approach and land in order to avoid a go-around, and he called for an increase of the airplane's "V-ref" airspeed by 5 knots to compensate for a potential "split flap." The approach was continued to a safe landing. The captain stated that the first officer "had to use a little bit of right rudder" during the approach and landing, and the control yokes were "about 3/4 of the way over" to the left. The captain stated that he was "thankful we were wings level" at the time of the incident, and he estimated that the incident occurred as the airplane was flying between 500 and 700 feet above the ground. He stated that no altitude was lost, and that the cockpit resource management during the incident was "excellent." The captain stated that the flap control was not moved after the landing and during the taxi to the gate. After parking the airplane, he egressed and inspected the airplane. He noticed that the right wing flap sections were "split" into an asymmetry; the inboard flap was lower than the outboard flap. He also stated that the inboard portion of the right outboard flap "came loose and cocked the aileron." He also observed that the "upper surface of the right outboard flap was in contact with the lower surface of the right aileron, " and the aileron was "hard to move" by hand. According to the first officer, who was the flying pilot at the time of the occurrence, the incident occurred when the airplane was about 500 feet above the ground, 140 knots, wings level, the "landing assured," and the flaps in transit from 17 degrees to the next and final setting at 35 degrees. The first officer stated that "a couple of seconds" after the flaps were beginning to lower, the "airplane started to abruptly turn to the right...at least 15 to 20 degrees." The first officer stated that he immediately "countered to the left" by moving the control yoke, at which time he "felt a pop." He stated that he "popped through and broke something loose" and "both hands were on the yoke at the time I did it." He also stated that the force needed to counter the right roll surprised him, and he reacted quickly. He first thought it may have been wake turbulence. He stated that if the situation would have occurred with a pilot who had less physical strength and size than himself, the abrupt roll "would have been a shock." The first officer stated that he needed to keep in left control yoke in order to keep the wings level. He stated that the control yoke was at "almost full deflection" to the left, and it was pointing toward the "nine o'clock or ten o'clock" position. He also needed to put in "some rudder to compensate." He stated that the situation would have been much more hazardous if the abrupt roll and jamming would have occurred while the airplane was in a right bank. The first pilot was the captain and non-flying pilot of the incident flight. He was an airline transport pilot and had been flying for Mesa Airlines since June, 1994. The first pilot reported that he had about 3,100 hours of total flight time, including 2,850 hours in type. The second pilot was the first officer and flying pilot of the incident flight. He was also an airline transport pilot and had been flying for Mesa Airlines since July, 1996. He reported that he had about 1,800 total flight hours, including 330 hours in type. The aircraft, a Beech model 1900D, was manufactured in 1993. Two flap sections are mounted on each wing of the Beech 1900D and travel in unison via an electric motor. The outboard edge of the outboard flaps are located next to the inboard edge of the ailerons. The inboard flap sections are larger than the outboard flap sections and incorporate sealed roller bearings to facilitate their movement. The outboard flaps sections incorporate non-sealed roller bearings. According to the Raytheon Aircraft Company, an upgraded design for sealed roller bearings had been company-approved prior to the Seattle incident, but the vendor was unable to provide the bearings at the time of the design approval. The flap system design for the aircraft incorporates asymmetric limit switches which are supposed to stop the flap movement when such a condition first develops. Two linkages are mounted between the flap sections and are connected to a device which incorporates a roller cam. Should one flap section continue moving while the other does not, the relative motion between the flap sections -- and thus the linkages -- causes the roller cam to ride up onto the flap limit safety switch, thus deactivating electrical power to the flap motor. The device is designed to activate should one flap section move more than 3 to 6 degrees relative to the other section, assuming the flaps remain properly connected to their attach brackets. An examination of maintenance records (excerpts attached) revealed that the Seattle incident airplane had accumulated 10,642 cycles. It received a routine inspection five days prior to the incident. The routine inspection did not include specific instructions of a detailed inspection of the flaps. The airplane also received a "detailed inspection" on November 2, 1996; according to the inspection checklist, no discrepancies were found on the flap skin or structure and the inspection did not include specific instructions to inspect the flap bracket or roller bearings. It was also determined that the airline had not been utilizing the most recent Beech 1900D maintenance manuals at the time of the incident. Because Mesa Airlines did not immediately report the incident to the Safety Board, the flap was removed and replaced by the airline prior to an accurate measurement of the flap asymmetry by the Safety Board. One day following the incident, the Safety Board examined the affected flap and its remaining roller bearings after it had been removed. The examination revealed that the aft roller bearing of the right inboard flap track had seized, and its associated flap bracket was torn. The outboard edge of the right outboard flap, which is positioned next to the inboard edge of the right aileron, exhibited evidence of gouging. All four forward and aft roller bearing assemblies and the entire right outboard flap was sent to the Raytheon Aircraft Company (formerly the Beech Aircraft Company) for laboratory evaluation under the personal supervision of the Safety Board. It was noted during the laboratory evaluation that the aft outboard outer roller was found loose on the bearing and could be shifted axially on the bearing outer ring. The bearing assembly consists of a standard roller element bearing pressed into a cylindrical outer roller with a flange on one end. The installation provides clearance at each end of the outer roller when the bearing assembly is positioned between the two supporting mounting bracket skins with a mounting bolt installed through the hole in the center ring of the bearing. A circular piece of metal that resembled a washer and was similar in dimension to the diameter of the outer roller bearing flange was found on the roller element shaft. The metal was from the material that comprised the skin of the flap track mounting bracket. According to personnel at Raytheon, the bearing outer roller had shifted on the roller element bearing and allowed the outer roller flange to wear against the side of the mounting bracket skin until the assembly pulled out of the bracket. The examination further revealed that due to damage that occurred to the bearing outer roller during jamming and binding of the bearing, the dimensions and degree of interference fit between the outer ring of the bearing and the bearing outer roller could not be determined. An examination of the other three roller bearing assemblies from the right outboard flap revealed that two of them were covered in a fine black debris, exhibited evidence of erosion, and required replacement. Data collected from the Safety Board's accident/incident data base and Mesa Airlines revealed that this incident is the third known incident involving similar circumstances of the Beech 1900 aircraft series. The first reported incident occurred on March 17, 1995, in Los Angeles, and is documented in NTSB case number LAX 95IA142. In the Los Angeles incident, the flying pilot noted in his interview that the roll onset took him completely by surprise and required full application of opposite aileron at "considerable wheel force" to counter the roll. He also stated that he might have lost control of the aircraft if the wind conditions had been unfavorable or the aircraft had been at a slower airspeed. An examination of the airplane's left outboard wing flap revealed a failure mode similar to the Seattle incident. As a result of Raytheon's review of its maintenance manual after the Los Angeles incident, an addition of an action was published to "Remove flaps and inspect flap roller brackets, rollers, bearings and attachment hardware for wear every 10,000 cycles or 5 years, whichever occurs first." Raytheon also recommended to inspect, without removing the flap, the "... roller bearings and attachment hardware," during the first 200-hour-interval detailed inspection and every 1,200 hours thereafter. These changes were made to the Beech 1900D maintenance manual in February 1996, without any dedicated communique' to announce the changes to the field. Similar changes were made to the Beech 1900/1900C aircraft maintenance manual and were published on April 26, 1996. No action to mandate the actions was made by the FAA. The second incident involving a flap asymmetry with a Beech 1900 derivative occurred on June 7, 1996. United Express flight 906, a Beech 1900D, en route from Albuquerque, New Mexico, with 13 passengers aboard, was approaching to land on runway 14R at Carlsbad, New Mexico. About 1 to 1-1/2 miles from the runway, and at an altitude of about 1,500 feet above the ground, the captain selected flaps 35. The captain reported that immediately after selecting the flaps, "the yoke twisted out of my hands to the left" and he had to induce "2 or 3 hard jolts to the right" before he "popped through" and the "yoke seemed to break free and return to zero angle of bank." He estimated that the airplane had reached a bank angle of about 40 degrees before he was able to level the wings. During the remainder of the approach, the control wheel could only move "about one inch of travel" when rotated in either direction, and that he had "very little aileron control." He continued the approach, and the landing was uneventful. After parking the airplane, he inspected the left flap and noticed that it was "...tilted up on the inboard side, tilted down on the outboard side" and "...the edges of the aileron and flap metal were ripped." The captain stated that the weather was VFR and calm winds prevailed at the time of the incident. He stated that a pilot with less physical strength would have had greater difficulty in controlling the airplane given the same incident. He also stated that he now only lowers the flaps when the airplane is in a wings-level attitude. The Safety Board was never notified of the incident and no investigation was conducted at that time. The Safety Board learned of the incident during the investigation of the most recent incident that occurred in Seattle six months later. A subsequent examination of the maintenance records (excerpts attached) revealed that the aileron needed to be replaced, and that the airplane had accumulated about 9,000 cycles at the time of the occurrence. The records also indicated that Mesa's FAA-approved maintenance program did not incorporate the changes that were published by Raytheon as a result of the Los Angeles incident. A review of Raytheon's Product Improvement Committee (PIC) database, which tracks field reports of airframe problems, cited 22 Beech 1900D occurrences of worn aft roller bearings and damaged flap brackets. None of the 22 airplanes had accumulated more than 6,000 flight hours. One entry stated "Pilot reported right flap asymmetry on final approach. Troubleshot. Found right [outboard] flap [inboard] aft roller ripped from flap causing it to jam. Removed and replaced the flap and roller. Checks now Okay." The Beech 1900D, serial number 199, had accumulated 5,191 hours at the time of the occurrence. The Safety Board also collected information from the FAA's Service Difficulty Report (SDR) database. One SDR was found which documented that a Beech 1900D "right outboard flap, inboard mount bracket worn. Removed and replaced flap assembly." The entry was dated June 25, 1993, and the airplane had accumulated 1,570 hours at the time of the occurrence. As a result of the aforementioned incidents, the Raytheon Aircraft Company (RAC) issued Safety Communique No. 137 (attached) entitled "Inspection of the Outboard Flap Attachment Brackets and Aft Roller Bearings," in May of 1997. According to the Communique: RAC has received three reports of outboard flaps becoming detached from the flap aft roller bearings at the outboard flap inboard flap track. Detachment of the outboard flap from this roller bearing could result in a flap asymmetric condition and the outboard end of the flap coming into contact with the aileron and inhibiting the travel of the aileron. The detachment of this flap aft roller bearing from the outboard flap was the result of the outer flange element of the roller bearing becoming repositioned on the bearing. When this happened, the outer flange element came into contact with the attachment bracket and eventually wore through the bracket, allowing the outboard flap to detach from the aft roller bearing. The following inspection shall be accomplished as soon as possible after the receipt of this Safety Communique' or in accordance with the following schedule on all applicable Beech 1900 Series Airliners and Model C-12J airplane with more than 2,600 cycles on the flaps. Aircraft with over 10,000 cycles on the flaps: Inspect within 30 days Aircraft with 5,001 to 10,000 cycles on the flaps: Inspect within 60 days Aircraft with 2,600 to 5,000 cycles on the flaps: Inspect within 90 days