Clandestino;
Thanks for keeping the dialogue on this drill alive. The disagreement on how to apply the drill in force at the time.
Originally Posted by Clandestino in post #3167
Emergency procedures are not written by company lawyers, they are written by test pilots.
You're probably right but I'm not a test pilot and don't actually know how drills and checklists obtain the form they do. I do know that the UAS drill has many iterations throughout its history since it was first instituted after the Birgenair and Aeroperu crashes.
When the Airbus FCOM Bulletin and the UAS drill first emerged, a pitot (or static) failure was assumed to have occurred at takeoff as the two accidents had. Of course, it is an emergency under those circumstances and the guidance in the memorized drill was appropriate.
An ADR Disagree event occurred on an A330 as early as 1996. The characteristics were pitot icing, an undesired stall warning and a latching of Alternate law for the rest of the flight. I believe these events drove some of the changes which we see today regarding the stall warning, (inhibition above M0.866), the (10sec?) delay in latching Alternate Law, etc. Airbus issued Bulletin #11 dated October, 1997.
I don't know if an Unreliable Airspeed Drill/Checklist was in place at that time or not. The earliest drill I know of is November of 2002; there may be examples prior to this date.
Both FCOM Bulletin 11 and the UAS drill at the time, (found in 1.02.34 - Navigation) stated that the memorized pitch attitude and thrust setting were to be flown. The UAS drill qualified this by stating under "How to Apply This Procedure":
" - if the wrong speed or altitude information does not affect the safe conduct of the flight, first apply the ADR Check procedure to identify the faulty ADR(s) and switch it (them OFF. If necessary, enter the unreliable speed procedure, or severe turbulence table (if in cruise), to set the pitch and thrust corresponding to the current flight phase.
...
- if the safe conduct of the flight is affected (all the speed indications are unreliable, ro the wrong speed indication cannot not [sic] be positively identified):
- immediately apply the memory items: AP/FD/ATHR OFF, and fly the memory pitch - thrust settings";
- Then, once stabilzed, refer to the QRH in order to determine the pitch and thrust settings required by the current flight phase;
... " etc
There is no guidance as to what the meaning of "stabilized" is in this context. Does it mean, "stabilized in the (resulting) climb"?, or does it mean, "stabilized in level flight"?, because setting a pitch attitude of 5deg in cruise is going to result in a strong climb.
In my own experiment* the pitch to 5deg causes an initial climb rate of about 4000fpm with a commensurate decrease in airspeed even with thrust set in the CLB detent. It takes about two minutes to reach FL400 by which time the airspeed is around 200kts.
These numbers are nominal, and are in the same ballpark as those seen in the AF447 data.
Two minutes is not a long time as we know; by the time the QRH is brought out and the page found and numbers read, the airplane is already a long way from stable, level flight and is climbing while the energy level is reducing.
Later UAS drills are not materially different. However, two flight crew training manuals I am aware of states,
"A330/A340 FLIGHT CREW TRAINING MANUAL - ABNORMAL OPERATIONS, NAVIGATION
Rev.No.01
1st February 2007
MEMORY ITEMS
The flight crew applies the memory items, if the safe conduct of the flight is affected. The memory items allow to rapidly establish safe flight conditions in all phases of flight and in all aircraft configurations (weight and slats/flaps).
The flight crew must apply the memory items, if they have a doubt on their ability to safely fly the aircraft in the short term with the current parameters, ie:
• The flight crew has lost situation awareness, or
• The current pitch and thrust are not appropriate for the current flight conditions, or
• The aircraft has an unexpected flight path for the current flight conditions.
When the PF has stabilized the target pitch and thrust values, the flight crew applies the QRH procedure to level off and troubleshoot the problem. The flight crew must apply the QRH procedure without delay, because flying with the memory pitch/thrust values for an extended period of time can lead to exceed the aircraft speed limits.
Note: The flight crew must respect the STALL warning."
So, although the drill/checklist state that setting 5deg of pitch "protects" the airplane from stall, I see the opposite, and I can't see a test pilot actually thinking this was a better solution than remaining level while the QRH was brought out.
One can certainly re-establish stable, level flight by immediately flying the QRH pitch and thrust settings, but what has been lost is the situational awareness of what one's speed is. We know that at cruise flight levels there is not a lot of reserve power. So instead of staying level and keeping the pitch and thrust that was "successful" immediately prior to the loss of airspeed information, the crew loses that awareness by climbing and losing energy, which must be regained.
I would argue that it is not the "Is the safety of flight impacted?" that should be the qualifying condition, but the flight phase and the altitude at which the failure occurs. That removes the possibly-subjective assessment of whether one's aircraft is in jeopardy or immediate danger, and instead outlines those conditions in which the aircraft either is, or is not in immediate danger, which is predicated on the Thrust Reduction Altitudes and the Takeoff flight phase.
At cruise flight phase the airplane isn't in immediate danger if the airspeed information is lost. One has pitch, altitude, VSI and thrust indications from which stable flight can be maintained while the pitot's and ADRs sort themselves out, or if they don't, one has a stable airplane in level flight from which to troubleshoot the ADRs.
This emminently satisfies the first rule of aviation - "Aviate". Once the aircraft is under full control, then carry on with the drills, by first announcing them so that everyone knows what you're doing and can monitor and provide assistance. One continues to Navigate which means ensuring headings are suitable while the abnormals are being looked after, and one continues to Communicate which means using standard calls to announce the abnormal or the drill and, where an action is irreversible, (engine shutdown, for example), confirmation from the other crew member, (PF) before taking action.
To address Mountain Bear's earlier point, none of this is the result of "hindsight bias" or second-guessing the crew. I am well aware of the issue and have written about it on PPRuNe. I don't claim immunity from the bias but what is the boundary between hindsight, and examining what occurred and is known in the data? It isn't easy to establish. We have to start somewhere and then remain cautious. While we may not know what the PF saw on his PFD or elsewhere, there is nothing which should have prevented the crew from carrying out these SOPs, which are also designed by test pilots and manufacturers as well as individual airlines. The question will be, Why were these actions not carried out in the manner trained?, which is a reasonable question to ask for the Human Factors Group.
I am glad that the discussion on this continues as I think it is a material factor in this accident. We do not know what was displayed on the PF's PFD, (although we know that the BEA is pursuing this possibly through the QAR) and we do not know what UAS training was provided or what reactions where recalled.
But the fact that there is some disagreement in how this should be done, means that at least a discussion is warranted and useful. I have yet to see or hear a good reason why flight with pitch and thrust settings which existed prior to the loss of airspeed data should not be maintained and instead the aircraft pitch set, by rote, to 5deg and then without delay, (as per the FCTM remarks), the QRH Procedure to level off and troubleshoot the problem" must be done to avoid exceeding aircraft speed limits, (which we must assume means both high and low speed boundaries).
Thanks again, Clandestino. Agreement isn't the goal, but thinking about the way this drill and checklist is designed and intended to be executed, is.