Originally Posted by
takata
In auto mode, the amount of heat provided by the heaters is modulated following what the "probes" themselves are sensing. When anti-ice is selected ON, full heating is provided, whatever the "probes" are sensing. It makes sense if you are considering that the "probes" are freezing because the heaters were not delivering the correct amount of heat at the first place.
In fact, it seems that those probes freezing at cruise are not related to temperature (really) changing but to those specific (very small) ice particles melting when such conditions are encountered (mostly in tropical weather). Hence, TAT will increase very fast and reach a level close to 0°C (iced particles temp melted on the TAT probes). The drawback is that you can not fly with anti-ice ON all the time as it can also damage all the systems by overheating the probes. Hence, it should be selected on purpose.
I'm sure I will be corrected if wrong, but I believe the issue with pitot tube heating is only when on the ground - no problem dissipating 300+ W when exposed to cold flight-speed air, but they'd glow red hot if left on on the ground! So the auto control will be very simple - as all safety critical systems should - to be easily testable, and surely a cut-off temperature would be near +40 degC than near ice-point?
I am not a big fan of the super-cooled icing theory, I suspect the current issue still subject to research is due to the way certain distributions of micro-fine ice particles (i.e. what makes up different clouds) pack acting as an insulator so there is poor heat transfer from the heated metal to the ice blockage. If it is the type of cloud - then I don't like the odds when identical probes pointing into the same one, that any 2 out 3 may be adversely affected in exactly the same way. If I was responsible for the pitot-tube detail design I would pay most attention to the drain hole and make sure that is always kept clear - meaning the most likely failure mode would be for value to fall to zero - easily detectable. Early Thales pitot tubes had poorly manufactured drain holes, burrs left which encouraged solid deposits and blocking, but sure this not now an issue.
As discussed many times drain hole blocking on its own would cause a small over-read, both ram and drain blocked could lock an old speed value in, but may be difficult to detect because speed reading still live - varies with static pressure reading/altitude. Worst time for this to happen would be during a commanded climb or speed change.
The pitots may only all disagree only when they melt out differently, so what was happening just before 02:10?