Investigators have disclosed that a 100t weight error led the tail-strike protection mechanism to activate on an Air France Boeing 777F, spurring the crew to command full take-off thrust.
Take-off parameters for the aircraft (F-GUOC) had been calculated using a weight of 243t rather than the actual figure of 343t, says French investigation authority BEA in its findings from the 22 May 2015 incident.
This gross error meant the calculated rotation speed of 152kt, with flap position 5, was far below the 175kt and flap 15 required.
Five seconds after the 777F rotated at 154kt from runway 26R at Paris Charles de Gaulle, its tail-strike protection system activated. The system used its maximum authority to stabilise the aircraft’s pitch at 9°.
The crew commanded maximum thrust some 8s after the protection system engaged, after the aircraft had managed to climb to 16ft. It was travelling at 189kt.
BEA says the 777F passed the opposite-direction threshold at a height of 172ft. There was no stall alarm and the stick-shaker did not activate.
Four cockpit crew had been on board the aircraft, which was bound for Mexico City. As it reached 20,000ft one of the supplementary pilots took over from the first officer, who had been flying and been startled by the incident.
While the cockpit-voice recording was unavailable to the inquiry, the investigators state that the crew discussed returning to the airport but ultimately decided to proceed with the flight.
BEA says the crew had been “aware” that the aircraft would be heavy, with a take-off weight close to the maximum of 347.4t owing to the payload and fuel.
But the inquiry says the first officer, who added the zero-fuel weight and the expected fuel load, made a “miscalculation” of 100t.
The incorrect figure of 243t was entered into the on-board performance tool, part of the electronic flightbag.
Investigators found that the captain also calculated the take-off weight and correctly determined the figure of 343t, but then entered 243t into the tool.
Owing to the absence of the cockpit-voice recording, says BEA, the inquiry could not determine whether the pilots’ individual errors were made independently or whether one of them was influenced by, for example, the verbalising of an erroneous figure by the other.
Cross-checking of the take-off parameters from the performance tool showed consistency between the two pilots’ data entries.
Since a large proportion of 777 and 777F departures at Air France were conducted with a flaps 5 setting, this configuration arising from the performance calculation did not arouse suspicion from the crew.
The figures from the tool were then fed into the flight-management system.
When the crew made a final calculation, based on the definitive load sheet, another miscalculation resulted in a take-off weight of 241.5t instead of 341.5t – with the result that this appeared consistent with the previous incorrect figure of 243t.
The computed configuration of the aircraft surprised one of the supplementary pilots, says BEA, but he did not air his doubts.
BEA adds that the procedures used did not provide for a coarse means of error detection, and that while a message from the flight-management system – referring to the unavailability of V-speeds – might have signalled an inconsistency, the message content was inadequate to alert the pilots.
When the flight-management system cannot calculate these reference speeds, says BEA, the crew is not warned about the loss of protection against entering speeds below the minimums normally computed by the system.