On Monday, June 10, 2013 3:05:24 AM UTC-7, Tom Gardner wrote:

Consider a refrigerator in which a bowl is in contact with
the cooling coils, and the coils have a layer of frost.

Put bowls of hot and cold water onto the frosty coils.
Hot bowl melts the frost to produce a puddle of water,
but the cold bowl does not.

Water re-freezes so that the hot bowl is in much better
thermal contact with the cold coils, so heat is removed
faster and the hot water freezes sooner.


Seems like an unlikely scenario. I've done it and the cold water melts the frost too because it takes a lot of energy to freeze water (cold or hot) which comes out of the frost. If the cold water was so cold that it was partially frozen, then maybe the frost wouldn't melt, but then it would be so far ahead of the warm water in terms of the enthalpy of fusion that the hot water would never catch up.

These thought experiments make a lot of assumptions that turn out not to be true - usually because the magnitude of the various effects at play are poorly understood by the people doing the thinking.

Quite right it's not applicable to the question from the OP.

BTW - in my empirical study, I was surprised how solid the ice became well prior to all the water being entirely frozen.

9B