PPRuNe Forums - View Single Post - AF 447 Search to resume
View Single Post
Old 12th Apr 2011, 18:29
  #3379 (permalink)  
Lonewolf_50
 
Join Date: Aug 2009
Location: Texas
Age: 64
Posts: 7,283
Received 461 Likes on 289 Posts
Svarin

In an attempt to follow the bouncing ball

From your previous

02:12:10 FLAG ON CPT PFD FPV
02:12:16 FLAG ON F/O PFD FPV

This was preceded by (among other things)

02:11:49 PROBE-PITOT 1X2 / 2X3 / 1X3 (9DA), HARD
02:11:55 WRG:ADIRU1 BUS ADR1-2 TO FCPC2, HARD

With the six second stand-in-line-for-next-message protocol, these are either roughly simultaneous events, or events in immediate sequence.
(Is that the point you were trying to make?)

Around ten seconds later, both pilots had tried and failed to get the FPV restored, and failed to do so. (In my mind's eye, I see the conversation "Mine doesn't work, does yours?" "No" and so they try to troubleshoot via other means ... and must resort to other indications for basic flight information)

The above followed a bit later by

02:13:45 F/CTL PRIM1 FAULT
02:13:51 F/CTL SEC1 FAULT

First, this describes a simultaneous failure of PRIM1 and SEC1, which greatly reduces the likelyhood of a manual shutdown or reset.
Second, this has a fascinating consequence, which can only be found when looking carefully at the flight controls reconfiguration schematic in the FCOM: PRIM2 becomes Master of pitch, through both elevator halves and the THS (trim) because PRIM1 and SEC1 have given up.
Third, it is interesting to relate this mastery over pitch by PRIM2 with another event affecting the same computer in a strange way :

02:11:55 WRG:ADIRU1 BUS ADR1-2 TO FCPC2, HARD
(FCPC2 is PRIM2)
What the heck does this message mean?

(snip your most excellent exposition on that question)
From what you presented, it appears that PRIM 2 had already failed before it was called upon to act as the back up, as you described above.
How strange is it that this PRIM2 computer would find itself in the middle of an erroneous air data situation while at the same time suffering a connection failure with one of its three sources ? How would it react in this condition? How would it compute its flight controls commands ? Has this been tested or even considered during design ?
It appears that is has now been encountered once in the field, which perhaps provided the two gents on the flight deck the dubious honor of becoming test pilots ... without a test plan, with a load of pax, and with no briefing for the data points to be collected.
Lonewolf_50 is online now