AF447

[DJ77]

So its not icing, but still the only known problem with these probes (as far as I know) is the drainage, and apparently its not easy to solve.

HN39,

It's not icing at the drain end of the Pitot but eventually, I think the front end ice up.

The following is a surmise entirely (I'm not an engineer) but it is the way I think things happen.

When drain flow can no longer evacuate all the incoming water, the tube starts filling. At this stage, total air pressure is still transmitted through liquid water with little error and possible instability. However, while the tube is filling up, the heat flow from the de-icing elements is modified by the increasing water volume and becomes gradually unable to efficiently deice the forward tip of the tube. When the tube is clogged, pressure falls rapidly through the drain at the rear and air speed indication follows. Since the liquid water is still heated it could start boiling close to the drain and cause intermittent overspeed alarms.