Chronology Current Month Current Thread Current Date
[Year List] [Month List (current year)] [Date Index] [Thread Index] [Thread Prev] [Thread Next] [Date Prev] [Date Next]

Spikes growing from ice cubes




Anyone here heard any more about upside-down icicles growing in the
icecube tray? I"ve been kicking around a theory, and today I saw the
effect for the first time. The departmental fridge got a weekend
cleaning, and today there were two spikes in the icecube trays; one over
1in tall. I saw some evidence in support of my idea on how they form.

It goes like this: place an icecube tray in a room-temperature
refridgerator and turn on the power. The cooling coils will chill the
bottom of the tray and ice will form at the bottom of each well in the
tray.

As the air temperature drops, the ice in the cubical well grows inwards
and upwards, driving a hemisphere of water ahead of it. The ice expands
as it forms, and the water expands as it cools, and so the small area of
"open water" at the top of the cube develops a positive meniscus. The
edges of the meniscus freeze. the liquid/solid boundary down within the
cube advances, the water and ice expand, and the positive meniscus is
again pushed upward past the frozen edge. The shrinking pocket within the
main body of the cube acts as a "squeeze bulb" and slowly pushes water out
the top. This only occurs because water expands upon freezing, of course.
A tube of frozen water develops, a tube with a liquid core and a liquid
drop at its tip.

As the pocket of liquid shrinks, the expansion of its volume becomes less
and less, and the edge of the meniscus contracts faster than the bead of
water being pushed upwards, until... HEY! I just read this last month in
SCI NEWS: liquid droplets develo p sharp peaks! Which is exactly what
happens with the ice spine; the ice closes as the liquid volume shrinks
and stops expanding, and this ends with a singularity, a sharp point. An
ice spike is just a special case of the sharp point which develops on a
droplet which freezes bottom-up.

_
/ \ Liquid droplet
|| || at the tip.
|| ||
|| ||
|| ||
|| ||
|| ||
_______// \\_______
\ / \ /
\ | | /
\ \______/ / Ice cube with fluid pocket
\ / and "spike" being constructed.
\ /
\ /
\______________/


|
||
||
| |
| |
| |
| |
| |
| |
| |
| |
| |
_______/ \_______
\ /
\ /
\ / Ice cube with completed "spike"
\ / and all liquid solidified.
\ /
\ /
\______________/


So, the key to reproducing the effect would be to force the cube to freeze
bottom-first, perhaps by placing warm water in a container upon a chilled
plate. Or perhaps a tiny bit of gelatine could be used to prevent
convection and aid the bottom-first freezing. The air temperature
probably should be below 0C, but if it is TOO low it will cause the open
water to freeze over prematurely.

The classic report of this effect is when a refridgerator suffers a long
power failure. The freezer section warms up and the ice melts. When
power is restored, the icecube trays are against a chilled plate, so
freezing can take place starting with the bottom.

There are reports of ice spikes which lean sideways at extreme angles.
This might be caused by air motion disturbing the rate of freezing at one
side of the meniscus? Also, ice spikes are frequently prisms rather than
perfect cylinders. Their cross-sectional shape might be programmed by the
shape of the open patch of water at the top of the cube at the onset of
spike growth?

......................uuuu / oo \ uuuu........,.............................
William Beaty voice:206-781-3320 bbs:206-789-0775 cserv:71241,3623
EE/Programmer/Science exhibit designer http://www.eskimo.com/~billb/
Seattle, WA 98117 billb@eskimo.com SCIENCE HOBBYIST web page