An A330 aeroplane, equipped with Angle of Attack (AoA) sensors with conic plates installed, recently experienced blockage of all sensors during climb, leading to autopilot disconnection and activation of the alpha protection (Alpha Prot) when Mach number was increased. Based on the results of the subsequent analysis, it is suspected that these conic plates may have contributed to the event. Investigations are on-going to determine what caused the blockage of these AoA sensors. Blockage of two or three AoA sensors at the same angle may cause the Alpha Prot of the normal law to activate. Under normal flight conditions (in normal law), if the Alpha Prot activates and Mach number increases, the flight control laws order a pitch down of the aeroplane that the flight crew may not be able to counteract with a side stick deflection, even in the full backward position. This condition, if not corrected, could result in reduced control of the aeroplane. condition on Airbus A330 and A340 aeroplanes. AoA conic plates of similar design are also installed on A320 family aeroplanes, and installation of these AoA sensor conic plates was required by EASA AD 2012-0236, making reference to Airbus SB A320-34-1521 for in-service modification. This requirement has now been removed with revision 1 of that AD. To address this condition on A320 family aeroplanes, Airbus developed an “AOA Blocked” emergency procedure, published as a temporary revision (TR) of the Airplane Flight Manual (AFM), to ensure that flight crews, in case of AoA sensors blockage, apply the applicable emergency procedure. For the reasons described above, this Emergency AD requires insertion into the AFM of the Airbus TR. This AD is considered to be an interim measure and further AD action may follow.
With multiple frozen/blocked AOA probes and an increasing mach no, the flyby wire system might trigger alphaprot mode. As the system thinks the aircraft is stalling the flybywire system induces a constant nose-down input to the aircraft. The speed indication however will be normal as the pitot-probes are functioning normal. The nose-down input is not fedback via the sidestick, nor on the screen. And pilot full nose-up input might not be able to counter the nose-down input of the alpha-prot system.
So the brilliance of engineering instructs us in a procedure to turn off 2 ADR's in order to revert the aircraft to Alternate law, in which no alpha prot exist. Off course first you have to realise what is going on, which could be difficult as there is no feedback from the system to the pilot.
But yes you read it correctly, the system is able to completely override pilot input despite being completely wrong. And the only way to recover from uncontrolable situation is to switch of ADR's.
At any time, if the aircraft goes to an unmanageable pitch down attitude despite continuous deflection of the sidestick in the full backward position (in case the flight crew missed the below symptoms or delayed the application of one of the below procedures): Keep one ADR ON. Turn OFF two ADRs.
And before anyone asks, yes I fly Airbus A320 family. (not 737 as my name might suggest)
Big topic which I do not wish to trivialise in the least.
However there are two things involved here the first being the question
How safe is (airbus) fly by wire?
and the second the event and all the technicalities that gave rise to the question.
I only feel competent to comment on the question and my response is 'safe enough'. That word 'enough' results from the simple fact of the number of people killed in Airbus accidents compared with those killed in other makes in current use. When I started in the industry wings fell of Electras and fuselages split on Comets in the cruise and that was clearly not safe enough. Today so many people fly so cheaply (and safely) all over the world that their seat on an airliner is just another commodity ruled by market forces. This means (whether we like it or not) that the large amounts of money needed to sgnificantly improve the current safety level cannot be justified commercially and so will not be found. You or I might well be prepared to pay double or treble for a safer seat but you or I are not the market.
It is highly debatable that ALT II (b) actually killed 447. Some could argue that misidentification of a stall situation is the result of many different inputs and if the aircraft control law was at fault then Air France wouldn't have got so upset when the safety recommendations were published.
The very nature of our operation in aviation is such that an almost unlimited combination of failures and events can occur, with the likely, unlikely and even remotely conceivable designed for and considered by Airbus.
This AoA vane has been shown not to work after operational experience; the problem has been identified quickly, OEBs have been issued and (as uncomfortable as it is to think our aircraft can behave in erratic ways) the truth is that we as pilots make many more mistakes. How about the incident involving mis-selection of the rudder trim switch instead of the cockpit door lock? What happened after that?
As an A320 pilot, I did not feel comfortable reading the new OEB and background information. However, we can only follow the advice given, and for the rest rely on our training and experience, as we do in any other situation. Our job has inherent risks which we manage. Better we know about it now.
Last edited by The African Dude; 19th Dec 2012 at 13:47.
It's interesting that Airbus advice is to turn off two ADRs, which immediately adds to confusion in the flight deck with only one air data reference available.
But that is preferable to seeing increasing nose-down pitch which you cannot correct in Normal Law...
Is there any reason why the FACs couldn't be switched off instead to drop the aircraft into alternate law?
Short answer: A330/340 do not have FACs.
Thinking of the A320....? 1. The ADR pb's are located together on the overhead panel, directly above the captain's head, whereas the FAC pb's are one either side of the same panel. Much easier and quicker to use the ADR pb's.
2. Switching off both FACs has more effect on the rudder - it leads to Mechanical Yaw instead of ALT.