Originally Posted by ttse
1)You've heard of sublimation but at your low elevations probably haven't witnessed it. It's a common process in Colorado and I would expect melting ice on the exterior of pitot's to sublimate. An unknown question becomes whether heated ice crystals within the pitot system sublimates or melts into water.
I have actually been looking into this after reading of a number of pitot icing issues where I thought supercooled liquid water could be ruled out. Meteorological survey reports were a major source of information in that research.
I have concluded that there must be a secondary process by which ice forms in pitot tubes at high altitudes where there is a very low temperature and a high density of ice crystals. My guess is that sublimation is involved in this process, but I can't be sure. I can see two hypotheses for this icing:
The first is that the ice crystals deposit but don't form a hard solid mass. Since the ice will have greater mass than the air it displaces, it will also have more inertia, and thus a partial obstruction of the pitot tube with loose ice crystals would dampen pressure changes (venturi effect). The problem with this hypothesis in my view is that the small ice crystals are small enough to be dispersed more evenly and even occasionaly ejected by pressure changes which occur in turbulance. Thus I have trouble imagining this process causing problems over ten or twenty minutes.
The second hypothesis is that the ice crystals, once deposited, may sublimate and condence, fusing them together into a hard mass (similar to the way wind-crust develops on ski slopes). An obstruction of this sort would also reduce pressure via the venturi effect, but would be less prone to being ejected from the tube by changes in pressure.
There have been enough reports of pitot tube icing in met. surveys of MCS's outside of strong convection (in one of the main cases I came across, in a DC-8 ;-) ) that I think we can understand that there is a process that the pitot tubes can ice up in the absence of liquid water. What that process is exactly may be harder to verify.
The absence of supercooled liquid water does not mean that pitot tubes were unlikely to be iced up in AF447. The parallels I have been able to find suggest that the pitot tubes would have almost certainly iced up.