Alcademics reader Andy L commented on the post about how to Make Clear Ice Balls Using a Thermos with an idea:
Fill the ice ball with water and set the thermos upside-down on top of it like a dunce cap. This way the ice ball mold would still be insulated on the top half: Would directional freezing allow it to freeze from the bottom-up?
Alas it seems not. I attempted to do this two ways:
- The ice ball on bottom, empty thermos upside-down on top.
- The ice ball on bottom, partially-filled thermos upside-down on top.
The theory with the second set-up is that the water inside the thermos would provide additional insulation in case the air in the thermos wasn't doing it.
The reveal. You can see that in the partially water-filled one on the left that the water remaining inside the thermos is frozen. On the right you can see where the water was pushed out the top hole as it froze.
In both set-ups, the cloudy part of the ice was a tornado-shaped column right up the center facing the hole (the sphere on the right is rotated sideways in the picture). So the center was the last part to freeze.
Most likely when the first water froze it floated to the top and plugged the hole.
Alas, it would have been great. We'll add this to the list of "experiments that didn't work" on the Index of Ice Experiments page.