TurbineD
Vasquez:
Quote: "...With a flight level temperature of -40 deg C suggested by the proximity sounding (and -36 deg C parcel temperatures) the A330 would be exposed mainly to frozen ice particles and perhaps graupel..."
and:
"...Quote: Tropical storm complexes identical to or stronger than this one have probably been crossed hundreds or thousands of times over the years by other flights without serious incident, including ascents and descents through critical icing zones in tropical showers. My original conclusion from June 2011 is still unchanged: turbulence and possibly icing creating an initial problem that led to a failure cascade..."
Here, he isolates icing as ascent and descent related.
Here, Vasquez suggests an exposure to frozen "ice particles" and "graupel".
There is no water available, and any ICE related problem would be a "packing" of the Pitot void by micronucleated ice. As Hazelnuts39 points out, a loss of power due to this phenomenon is known, particles sticking to engine parts, (including blades, I think.) Engines have anti icing.
For ICE to plug the Pitot at its orifice, ice particles would have to attach, or enter and pack, or melt and refreeze, to include a plug of the drain, which itself has a proclivity for corroded drain holes.
I believe this is the only way pitots could be plugged in the FL Vasquez is researching. This phenomenon could be eliminated (has been?) by BEA with a simple examination of leading edges of structure, inspecting for the typical abrasion of metal, composite, and painted surfaces caused by solid ice particles impacting at high speed.
That's it for me on ICE, what I think happened is Unusual attitudes caused a discrepancy in airflows into and past the separate Pitot Tubes, and causing sufficient data disruption to fault the AD(IRU). Coupled with a large turbulent flow that may have upset airflow rates, the sequence of events at a/p disconnect (and prior to) was likely the beginning of UAS. If turbulence related, there was a prior input that caused (perhaps) the a/p drop due its own mechanical limits, these limits serving to cause the drop and subsequent upset.
Again, the previous fifteen seconds to a/p drop, and the following 15 seconds seem to me to be the beginning of the accident.