Air France 747 Runway Departure at Montreal-Trudeau

Incident Overview and Aircraft Details On the day of the incident, an Air France Boeing 747-428, registration F-GITC, serial number 25344, operated flight AFR346 from Paris Charles de Gaulle International Airport to Montréal/Pierre Elliott Trudeau International Airport. The aircraft carried 491 passengers and 18 crew members. It departed at 1456 UTC and landed on Runway 24R at 2145 UTC. During the landing rollout, the aircraft veered to the right. The crew applied maximum braking in an attempt to stop the turn. The aircraft came to a complete stop at 2148 UTC, with the nose wheel landing gear 26 feet past the right edge of the runway and the main landing gear still on the runway. No injuries were reported, and passengers disembarked using transfer vehicles. A post-flight inspection found no damage and no system malfunction. ## Environmental and Operational Conditions Weather at landing included a few scattered clouds at 24,000 feet above ground level, visibility of 30 statute miles, and winds from 290°M at 5 knots. Investigators found no indication that weather played a role. Runway 24R is 11,000 feet long, 200 feet wide, and aligned 238°M. Inspections that day showed no irregularities, and the surface was dry. Land and Hold Short Operations (LAHSO) were in effect at the airport but not used for this landing, leaving the full runway length available. ## Crew Background and Flight History The flight crew was certified and qualified. The pilot-in-command had over 10,000 flight hours and had recently returned to the Boeing 744 as captain after serving as an Airbus 320 captain. His transition training included line training flights and a satisfactory line check completed on 29 June 2008. He had performed 18 flights since being released as a B744 captain, and this was his 10th landing as pilot flying. The first officer had approximately 6,000 flight hours and had served as a B744 first officer for six years. The crew had been on duty for about nine hours, with recent rest periods of two and six days, respectively. Fatigue was not indicated as a factor. ## Sequence of Events During Rollout The approach was a straight-in approach in visual meteorological conditions with light winds. At 2144 UTC, the aircraft crossed the threshold stabilized in a flaps 30 configuration, with ground spoilers armed and autobrake set to position three per Air France standard operating procedures. The touchdown occurred within the touchdown zone, ground spoilers extended automatically, and thrust reversers were selected. Longitudinal deceleration was constant, and heading remained within 1° of the runway heading with small rudder inputs. Idle reverse thrust was selected at 80 knots. At approximately 65 knots, the captain, acting as pilot flying, placed his hand on the nose wheel tiller, initiating a right turn. This was corrected with momentary left rudder input. The aircraft then continued along the runway centerline with a slight forward tiller deflection, meaning the nose wheels were turned to the right. At approximately 45 knots, as the thrust reversers were stowed and the autobrake disengaged, the aircraft veered to the right. The pilot flying used a combination of forward nose wheel tiller and full left rudder in an attempt to stop the turn, but the rate of turn increased rapidly. Maximum brake pressure was applied, and the aircraft stopped at 2148 UTC on a heading of 277°M, with the nose wheels 26 feet off the right side of the runway. ## Analysis of Crew Actions The initial rollout actions complied with company standard operating procedures. The pilot flying held the control column forward of neutral to maintain pressure on the nose wheel. A few seconds after thrust reversers reached idle reverse, the control column was released. One second later, at about 65 knots, the pilot flying voluntarily placed his hand on the tiller, possibly anticipating use of the next available runway exit, which was about 1,300 feet ahead on the left. Analysis of nose wheel skid marks indicated the tiller was initially pushed and maintained forward about 50°, corresponding to a right nose wheel deflection of about 25°. At that speed, rudder was still effective, and the pilot flying successfully used left rudder to overcome the right turn. However, this caused the nose wheels to lose lateral adherence, inducing a skid. The autobrake provided directional stability by applying even braking on all main gears, so rudder deflection was no longer required to maintain the centerline, and the pilot flying removed the left rudder correction. Skid marks straight down the centerline for the next 600 feet confirmed that the forward tiller input was maintained throughout deceleration without being noticed. The initial tiller input could not be fully explained, but the arm's angular position matched its position on the control column one second earlier. It is possible the pilot flying inadvertently pushed the tiller forward, thinking it was the control column. As speed decreased, the thrust reversers were stowed and the autobrake disengaged. The loss of directional stability from the autobrake, combined with the nose wheels regaining lateral adherence at 45 knots, caused the aircraft to veer right. In reaction, the pilot flying applied full left rudder and intended to apply left nose wheel tiller command. Instead, he pushed the tiller further forward and down, increasing right nose wheel deflection. Based on prior experience as a first officer in the right seat of the B744, this forward and down input on the right-hand tiller would have produced a left nose wheel deflection. It could also reflect an unintentional attempt to use an Airbus 320 tiller input, which is located lower and to the left compared to the B744 tiller position. At low speed, nose wheel steering inputs override rudder steering inputs, making the left rudder ineffective in stopping the turn. Had the tiller been released, it would have returned to the neutral position and the turn would have stopped. Boeing documentation advises against using the tiller above taxi speed. Although placing a hand on the tiller is not formally discouraged in existing procedures, doing so at a speed where it should not be used increases the risk of inadvertent activation. There was no pressing reason to turn the aircraft where the skid marks appeared, as no taxiways were nearby and the full runway length was available. This suggests the initial tiller application was an involuntary movement. When the aircraft later veered right, the pilot flying faced a time-critical situation. With his hand already on the tiller, he was likely influenced by habit recall from prior A320 experience or right-seat B744 experience. While rapidly attempting a left tiller input, he inadvertently pushed the already deflected tiller further down, increasing the right turn and causing the aircraft to depart the runway surface. ## Evidence Preservation and Recording Systems Airports Management personnel initiated work around the aircraft without prior consultation with Transportation Safety Board investigators to reopen the runway as soon as possible. Because the emergency response plan at the time did not include provisions for reportable incidents, it is likely they did not realize TSB approval was required. TSB regulations require preservation and protection of evidence relevant to reportable incidents. However, guidance for reportable incidents in the Aeronautics Manual did not include provisions for protecting the aircraft or occurrence site. This inconsistency can contribute to uncertainty about obligations to protect the site and preserve evidence after a reportable incident. The cockpit voice recorder recordings of the landing incident were overwritten, indicating that electrical power to the CVR remained on for a significant period after the event. The automatic shutdown feature was not installed on the event aircraft, so the CVR had to be manually deactivated by pulling the appropriate circuit breaker. After the aircraft stopped, the CVR circuit breaker was not located and pulled, and all relevant conversations were lost. Chapter 11 of the Air France Généralités Opérations manual, section 6, implies that preservation of CVR recordings is at the captain's discretion, while the checklist in section 8 requires deactivation of the CVR following a serious incident or accident. This inconsistency greatly increases the chance that critical CVR data will be lost despite regulatory obligations to preserve evidence after a reportable incident. This loss of CVR data is not unique to Air France and occurs frequently when aircraft are involved in reportable incidents, hampering investigators' ability to obtain a timely, complete, and accurate understanding of the event. The aircraft was not equipped with an image (video) recording device, so critical information that could have helped determine the events leading up to the occurrence may have been lost. The Transportation Safety Board laboratory report LP 113/2008 – DFDR/QAR/CVR was completed for this incident.