HISTORY OF FLIGHT On November 24, 1996, at 1450 eastern standard time (est), an Airbus A320-211, N310NW, operated as Northwest Airlines flight 211, experienced a rudder system malfunction approximately 50 feet above ground level (agl) while transitioning from approach to landing on runway 3L at Detroit Metropolitan Airport, Romulus, Michigan. The airplane was subsequently landed with the rudder stuck at zero deflection. While taxiing to the gate, the rudder system returned to normal. Instrument meteorological conditions prevailed at the time of the incident. The flight was being conducted as regular scheduled domestic air carrier service under 14 CFR Part 121. An IFR flight plan was on file. There were no reported injuries to the 6 crew and 141 passengers who deplaned normally at the gate. The flight originated at Fort Lauderdale, Florida, at 1130 est. In his written statement, the captain said that he "was at the controls, conducting a manually flown (autopilot off) ILS approach." He said that the weather conditions "were approximately 1,100 feet overcast with very light icing in the clouds. There was a right-quartering tailwind on final which subsided to a very light crosswind over the runway threshold. Due to this light crosswind, a slight application of rudder/wing low compensation was required for the flare and landing." When the captain attempted to use the rudder, he found that the rudder pedals "were locked in the neutral position." The captain used slight banking to keep the airplane on runway centerline through the touchdown. After touchdown, the captain used differential braking to stay on runway centerline. On reaching a ground speed of approximately 80 knots, the captain used nosewheel steering for directional control. During the landing roll, the captain asked the first officer to come on the rudder pedals with him. The first officer stated that he did not apply any force, but did notice that "the rudder pedals seemed stiff or possibly locked." After exiting the runway, the captain performed several autopilot disconnects using the autopilot disengage button on the control stick. "This had no effect other than to verify that there was truly no autopilot involvement." The captain then made several attempts to move the rudder pedals. "After about 15 seconds of this, the rudders became free and moved normally." AIRCRAFT INFORMATION Northwest Airlines Maintenance conducted a post-incident inspection at Romulus, Michigan, on November 25, 1996. First, a test of the autopilot disengagement system was conducted. The test revealed no anomalies in the autopilot system. Northwest Airlines Maintenance then examined all of the rudder system control components from the cockpit to the rudder control surfaces. No indications of mechanical anomalies were found. No signs of water or ice were found through the rudder cable inspection. The rudder attach points were cold soaked with dry ice. No binding was noted. The rudder autopilot artificial feel solenoid was removed, replaced and retained for further examination. METEOROLOGICAL INFORMATION The National Weather Service Weather Observation Facility at Detroit Metropolitan Airport's 1455 est observation was a 400 foot overcast ceiling, 3 miles visibility with drizzle and mist. The temperature was 33 degrees Fahrenheit. The dew point was 30 degrees Fahrenheit. The surface winds were 040 degrees magnetic at 7 knots. TESTS AND RESEARCH The flight data recorder was removed on November 26, 1996, and read out on December 6, 1996. A review of the flight data recorder readout showed that the autopilot was disengaged at the time of the incident. The data showed some movement in the crew's rudder pedal positions during the landing. The data showed little to no movement of the rudder surfaces. A review of all airworthiness directives and service bulletins with respect to the A-320 rudder system revealed an Airbus Industrie service bulletin dated March 21, 1992 entitled "Flight Controls-Rudder-Increase Radial Play of Lever Bearing in the Artificial Feel and Trim Unit." The service bulletin addressed conditions where the A-320 rudder artificial feel and trim unit did not disengage from the autopilot mode to normal pedal operating forces during approach and landing. "Investigations have shown that the radial play of the autopilot mode engagement/disengagement lever bearing together with low temperature could cause an increased operating force. In this case, the back driving force is not able to rotate the lever to get the autopilot mode disengaged." The service bulletin introduced a new modified lever with a larger radial play of the bearing to eliminate this problem. The A-320-211, N310NW, did not have this modification at the time of the incident. The rudder autopilot artificial feel unit was tested in the laboratories of Aerospatiale in Toulouse, France on January 27, 1997. A representative from the Bureau Enquetes-Accidents (BEA) was present to oversee the testing. The examination revealed that the "radial play of the autopilot mode engagement/ disengagement lever was not sufficient. This introduced a higher friction in the bearing of the lever, resulting in an increased operating force. In this case, the back driving force is not able to rotate the lever to get the autopilot mode disengaged. This results in increased pedal forces to move the rudder." ADDITIONAL INFORMATION Parties to the investigation were the Federal Aviation Administration, Northwest Airlines, Airbus Industrie of North America, and the Air Line Pilot's Association. The airplane was released and put back into service on November 25, 1996. The flight data recorder and rudder autopilot artificial feel solenoid were released and returned to Northwest Airlines.