Air Wisconsin Airlines Flight 3318

Following a normal touchdown, spoilers, reverse thrust, and wheel braking was used for the initial deceleration (about 5 knots/second), and the airplane tracked the centerline of the runway for about 9 seconds. About 1-2 seconds before the thrust reversers were stowed, the lateral acceleration began to increase (consistent with a right turning tendency). Subsequently, the magnetic heading began to increase as the airplane began a continuous right turn. The flight crew made two brief 4° to 5° left rudder inputs (each lasting 1 second or less) followed by increased wheel braking that provided more brake pressure on the left main landing gear wheels than on the right main landing gear wheels. The airplane continued to turn right, departed the runway surface into the grass alongside the runway, and came to rest on an adjacent taxiway. The airplane sustained minor impact damage to the left wingtip and a puncture in the left wing leading edge. The investigation was unable to determine what caused the airplane’s continuous right turn. The thrust reverser use was asymmetric with the peak right engine power about 12% higher than the peak left engine power, which would have produced a right-turn tendency. However, the amount of asymmetry was brief, not unusual, and unlikely to cause a continuous right turn. Once the reversers were stowed, any right-turn tendency due to asymmetrical reverse thrust would no longer be present. Three previous landings that involved the incident airplane revealed thrust asymmetries of 5% to 15% (favoring the right engine) during reverser use, and no loss of directional control ensued. For the incident landing, the right turn persisted well after the stowage of the thrust reversers. The flight crew’s two brief left rudder inputs and increased wheel braking on the left main wheels would have provided some left-turn tendency to counter the right turn; however, the right turn continued.

The reported weather conditions included a crosswind from the right, which could produce a right-turning tendency. However, the crosswind component was about 3 knots, which would have had a negligible effect on the airplane. The rudder pedals also control the nosewheel steering angle. A full rudder pedal input (25°) will move the nosewheel to a maximum of 5.3° left or right. Using the rudder pedals to sufficiently turn the nosewheel is particularly important as the airplane slows and the rudder becomes less effective for directional control. Had the flight crew made a larger and sustained left rudder pedal input, the rudder surface would have deflected to generate an opposing aerodynamic yawing moment and the nosewheel would have turned to the left (or turned farther so) for a longer period, both of which would likely have been more effective in countering the airplane’s turn to the right. The investigation found no indications of a nosewheel steering system failure or malfunction in the recorded data or the operator’s post-incident examination of the airplane. However, because neither the nosewheel steering angle nor tiller control were required/recorded parameters on the flight data recorder, a determination regarding the operation of the nosewheel steering system could not be made. The flight crew’s left differential braking (until the airplane departed the runway) would have provided some left-turning tendency to counter the right turn. Although differential braking is a technique that can be used to help steer the airplane, the operator’s flight crew manual did not reference the use of differential braking for a normal landing. The manual advised that rudder steering (via the pedals) should be used at all speeds. Thus, the flight crew should have used the rudder pedals which would have provided aerodynamic yaw via the rudder surface as well turning the nosewheel steering, to prevent the airplane from departing the runway.

Verbatim NTSB analysis and probable cause from the NTSB dataset