Ice lanterns -- what could go wrong?

Learning from mistakes

The great thing about ice lanterns is that many failures can be salvaged.  There aren't a lot of disasters in making ice luminaries.

Even the ice lanterns that crack during unmolding can often be pressed back together, because ice tends to weld in place with a little force applied.

Although this ice globe cracked (from just pouring
the water out of the bottom of the core after the upper
parts of the globe had cooled below the freezing point),
the cracks healed themselves without actually separating.

Here's an ice globe that cracked during
unmolding, with a large piece breaking off.
I simply pressed it back into place, and it
remained intact for days, with the cracked
side placed at the back, where it isn't noticed.

Ice lantern problems -- why they happen, and what you can do about them

What makes molds break?

One big issue is the expansion and contraction of water and ice.  Water contracts and gets denser as it cools, up to (or down to) a point.  That point is at about 40 degrees F (4 degrees C).  Below that temperature, ice crystals are beginning to form, and they are less dense than water.  At the freezing point, the solidifying water expands and takes up more space.  What does that mean for an ice lantern in a mold?  

Water shooting up from an ice lantern

Why is water shooting a foot up in the air from this ice lantern?  It's because of expansion during freezing.  The open top froze over, and ice began to form on the walls of the mold.  As the water turned to ice, it expanded.  This was happening during a very cold night, so the ice on the top was extra-thick and hard. To relieve the pressure, I melted a hole in the top and the water shot out, making a foot-high geyser.  

If the water in the core had continued to freeze and expand, it would eventually have broken the mold.

Why are some ice lanterns lumpy on the side that was open?

 Normally when the pressure on the water in the core builds up, some water finds a way up through the crust on top and wells up on the surface, where it either spreads out all over or freezes into lumps.  Differences in air temperature, wind speed, initial water temperature, vibration from traffic, and probably a host of other factors determine how this happens.

Using a constrained balloon inside the mold produces a flatter base and is a possible solution to this problem. (See how to use balloons inside other molds here.)

An electric drill with a quarter-inch (or so) bit can be used to drill a hole in the center of an ice lantern in a bucket as it starts to freeze over.  This releases the water in the core so that it doesn't build up as much pressure and so that it spreads more evenly across the surface.  The process may need to be repeated.  You can also poke a hole, but that tends to shatter the ice rather than make a nice hole.  The risk of breaking the mold is also decreased if the pressure is released.

The area on the upper right is raised above the
smooth surface of the rest of this ice lantern's base.

Why are some ice lanterns crystal clear and some have strings of bubbles all over?

This is only a problem if you insist on having one or the other.  The easiest solution is to be happy with either.  That's easy with ice lanterns, but if you were making metal castings, you would have to worry about this process, because you obviously wouldn't want your metal to be weakened by bubble chains. Knowing about bubble formation is also important for understanding ice stability in glaciers and icecaps.  A lot of research has been devoted to the conditions that promote or inhibit bubble development.  More bubbles form if there is more gas dissolved in the water.  The size and pattern of bubble formation is related to the speed of freezing. Chains form with slow freezing.  Bursts of tiny bubbles develop with rapid freezing.  

Many bubble trails.

Can you cut yourself on ice lanterns?

Yes.

Sharp edges do form on ice.

What is temperature shock and when can it be a problem?

Ice cracks if it is exposed to a sudden change in temperature.  This is particularly likely with water splashing onto ice that has chilled below the freezing point.  If you want to join two pieces of ice together, the best time to do it is when they have just been unmolded and have not cooled further.  Small amounts of cold water can be used to weld pieces of ice together, but be cautious and expect cracking if the water is too much warmer than the ice or the amount of water is too large.

Take another look at the ice globe from the beginning
of this page.  Those cracks were caused by a temperature
shock that surprised me.  I had taken a few minutes to put
the little "crown" of cocktail ice cubes around the opening,
leaving just a little cold water in the core while doing that.
When I tipped the lantern to pour out the cold water, the
wall of the globe had chilled enough in the sub-zero air
that the temperature shock caused it to crack.

This was an experiment with a small
dinosaur -- trying to freeze it inside an
ice globe -- with the result that the
globe simply exploded.

Wall thickness

It's hard to know when to unmold an ice lantern, and there are mistakes that affect wall strength and thickness.  

Avoid setting an ice lantern to freeze in a warm place (duh!).  That seems obvious, but it's sometimes a surprise how warm the sun can be, how much heat a house emits, and how hot the dryer exhaust is.  Don't put a dark balloon in the sun.  

Don't try to freeze ice lanterns at temperatures near 32 degrees F (zero C).  It takes too long, and the ice quality ends up being poor.  Try to aim for temperatures between zero and 23 F (minus 17 and minus 5 C) or plan on making adjustments.

A cold wind speeds the freezing process but may make it uneven.

When freezing ice lanterns at below-zero temperatures (below minus 17 C), check frequently to avoid freezing through and potentially cracking a mold.  Some molds are especially vulnerable to cracking;  the Arctic Ice Lantern is more vulnerable than many others -- watch it carefully.

Don't let snow accumulate on an ice globe or bucket lantern during freezing.  It can cause weak spots.  

Don't set ice lanterns too close together while freezing.  They can keep the air between them enough warmer to delay freezing on the sides that are near each other.  This can, however, be used to keep a spot open on the side of an ice flame or hanging drop;  freeze them in pairs that are touching.  The contact patch will stay open.  

Sitting on a warm or insulated surface will delay freezing.  If the bottom of the mold is the place that you want an opening to be formed, set it on snow or on ground that provides some insulation.  If the bottom needs to freeze solid, either pre-freeze a layer of ice in your bucket or make sure the bucket isn't resting on an insulating surface.  

If you have a big, important ice globe freezing, consider making some less critical ones at the same time.  You can unmold the cheap ones to get an idea whether it's been long enough for the important one. 

If the area that should be open or soft has frozen hard on your ice lantern when you go to unmold it, you can melt into the liquid core using a flat-bottomed metal bowl with hot water in it.  

This ice lantern had frozen almost solid,
leaving only a small liquid core.  It's
possible to melt a smooth-edged round
hole using a metal bowl with hot water.

As long as there is liquid water in the core, and for a short time after pouring it out, the ice lantern will be fairly resistant to thermal stress.  When the water is gone, the whole ice lantern will chill to the ambient temperature, and you have to be much more careful about splashing water on it.

See what happens when sunlight and warmth attack ice lanterns.


Return to the introductory page.