The world waits, perched collectively on the edge of its seat, for the outcome of the crash investigation into Air France flight 447 – in essence because this was one of the worst and certainly the most mysterious commercial passenger jet disaster in the history of commercial aviation.
The full, in-depth report won’t be ready until next year, however, the BEA did release a report based on the initial findings in response to speculation in the media over the findings from the flight recorders.
This seemed to corroborate with the general theory being put forward, but the BEA were adamant in pointing out that this only confirms the circumstances of the incident and not the cause of the crash. As is the case with incidents on this scale, no single fault – be it mechanical or human – will be solely responsible, it is almost always a result of many factors contributing with unfortunate consequences.
Paul Marks, senior technology correspondent for New Scientist explains some of the questions that still remain unanswered.
The mystery surrounding the crash of Air France flight 447 in June 2009 deepened last week when a preliminary report from air crash investigators revealed a double failure of the plane’s speed sensors – critical for fly-by-wire computer-assisted flight – followed by the flight crew inexplicably failing to maintain airspeed and deal with the stall that resulted,
The BEA says the plane hit turbulence at cruise altitude and then suffered temporary failures of two speed sensors – possibly due to high altitude icing. Because this meant an agreed speed reading could not be calculated – and speed is critical for maintaining aerodynamic lift – the flight computer switched off the autopilot and handed control to the pilots.
The nonflying pilot was heard confirming this disengagement had happened on the cockpit voice recorder saying: “so, we’ve lost the speeds,” and then “alternate law”. Fly-by-wire flight, where the plane is kept straight and level and at a safe lift-producing speed by a computer program (which has four backups) is managing flight in what is called “normal law”. This also ensures the plane can only bite into the airflow at a low, lift-producing “angle of attack”.
When the pilots are in charge the plane is said to be obeying “alternate law” – but here the angle of attack is not computer managed.
As control was handed over the stall warning sounded, yet the pilot flying the plane decided to climb. A stall usually requires the nose to be dipped downward to gain airspeed and regain the missing lift. Yet AF447 climbed inexorably to its ceiling of 38,000 feet at a very high angle of attack of more than 35 degrees. With the stall warning sounding, the nose remained tipped upward. The plane stalled, losing altitude at 11,000 feet per minute. It fell for 3 minutes 30 seconds when it hit the water nose up 16-degrees and belly first – just as the BEA had predicted from their analysis of floating wreckage.
This is the second time in recent memory that a flight has suffered from a nose up situation in a stall. Colgan Air lost an aircraft near Buffalo, New York, in February 2009 when drag caused by thick icing slowed a commuter flight so much it stalled. The pilot in that case also adopted a nose up attitude. It is not what pilots are trained to do.
Contributors to the online Professional Pilot’s Rumour Network – Pprune – are astonished that the AF447 pilots maintained a nose-up attitude. You can read their comments on (and either side of) this thread page. Some wonder if cockpits are being over-computerised leaving pilots baffled when it comes to manual flight in an emergency.
Others are worried about crew workload, coping with a blizzard of alarms in a thunderstorm in the dead of night. Mostly, they are so surprised the pilot kept the nose up in a stall – and for so long – that they think there has to be more to it: did other instruments fail?
The investigation continues.