Chris Scott said:
There seems to have been wide agreement here for a day or two that − in level flight, right at the beginning of this AF447 event − whichever ASIs were showing erroneous readings are more likely to have been over-reading the actual IAS/CAS, than under-reading it. Thus, if the crew were misled by the false readings, they might have slowed the aeroplane into a stall near the cruise altitude.
I can understand that, in the
absence of auto-throttles, that a progressively ice-accumulating pitot head would cause an “indicated” airspeed drop-off – and that the crew would be alerted and would be likely to intervene. Whether they’d interpret it as an actual loss of performance is another thing. With a subtle loss of indicated airspeed, the first inclination would be to start thinking engine icing and airframe ice excrescences gradually (and cumulatively) causing increased drag/perf loss.
However, with no disagreement (at least initially) between the three airspeed/mach indications, in an A330 the compensating autothrottles would simply incrementally pump more go-juice into the equation to maintain the stipulated cruise speed….. would they not? If the pilots had their fuel schematic (instead of engine parameters) up on the screen (often the case I believe, when fuel may be tight, especially in an A330) and with the non-moving Airbus throttles, this minor increase in fuel flow/decreased AoA and slightly lower pitch attitude would probably go unnoticed. That’s the nature of an insidious event such as symmetric ice crystal accumulation in three Thales pitot heads…… until the displayed airspeed split becomes significant enough to trigger a “disagree”. If they were at FL350 in layered CirroCU and CS, it’s not likely to have been more than pebbly turbulent along their radar-chosen best course between build-ups – so they’d have not been “on the qui vive” and expecting trouble (aka settled into mind-numbing cruise ennui, if you will).
The end result (see Belgique pg61/post #1208 at
link) of this quietly sinister acceleration would eventually be an aircraft headed, quite non-apparently, towards a Mach crit encounter. At, or even just prior to that juncture, either an airspeed disagree, pitch-trim disparity or (credibly) a mach-induced pitch-down may have caused the autopilot to click out – and then the crew would’ve been in a world of hurt as far as suddenly ascertaining
what was what amongst their confusing displayed performance criteria. That’s a recipe for snap decisions, cross-cockpit misunderstandings and a “sucked-in” and fatally incorrect PF choice…… particularly in an environment where very few pilots have handflown a degraded law in a ‘Bus in IMC at night whilst trying to troubleshoot. Bedlam might be an adequate description of the scenario.
A pilot well-drilled in believing his instruments might have trusted the still credible airspeed display on his PFD and unwittingly had a (follow-up) Mach encounter with throttles placed fully forward. I’m not sure to what mach number an A330 would’ve been tested, but I know a few experimental test-pilots who have a healthy regard for flying beyond Mmo/Vne – because there
“there be dragons”. It would be an easily achievable no-go area that can hold untold frights... in just this type of set-up. Airbus flight control protections are no longer in play when an aircraft is rolling, pitching and yawing as an aerodynamic direct result of a mach crit encounter. Having myself played those nose-low games in a quite robust jet trainer at great heights, I know that once you leave Mach country in total disarray and get below 20,000ft, things tend to sort themselves out. An airliner isn’t anywhere near as robust and could easily sustain significant structural damage in the first 5000 to 10,000ft plummet…. if the unusual attitude and pilot (or auto-trimmed to well out-of-trim) control response was extreme enough. Which brings me to the next query/point. What could induce a nose-down pitch into an Mmo encounter (which itself involves a reactionary nose-down pitching moment)?
.
During the quiet acceleration into/towards a mach crit encounter, what’s happening with the auto-trim and trimmable horizontal stabilizer (THS)? Is it getting conflicting inputs? (i.e. the computed trim-state from the ADIRU’s accepted and supposedly bona-fide indicated CAS airspeed - versus the actual dynamic pressures resulting from the actual airspeed). If the elevators were deflected and compensating to achieve an out-of-trim height-hold, what could’ve happened pitch-wise when the autopilot jumped out? Would the A330 autopilot have been opposing (i.e. holding) a nose-down pressure on its elevators due to undetected automation conflicts (or would it have been a pitch-up result at A/P disconnect?). Depends how the pitot/static data is being massaged into an autopilot interface - I guess.
We used to be able to declutch the autotrim in a PB20N autopilot and use the varicam trim (similar to THS) on the SP2H to squeeze a few extra knots (about 10 usually) for the cruise. It just achieved an “on the step” optimized attitude (was my diagnosis). It was a genuine increase validated by the navs, rather than an induced IAS position error). However if you forgot, and disengaged the autopilot without first removing the out-of-trim pressure being held by the elevators, you’d get a sudden nasty pitch-down – and some name-calling on the intercom. At coffin corner height in AF447, was this a likely contributing factor to a pursuant Loss of Control?
Any cogent input on the above suppositions?
.