Snap Shot: Amalfi Coast
Winter's Weirdest Cocktail Flavor Trend Has Bartenders Going Bananas

Clear Ice Cubes Using a Tray in a Cooler

Alcademics reader and ice nerd Mike Palmer has come up with an easy if not space-efficient way of making clear ice cubes using a silicone tray and a cooler.

The short answer is: poke holes in the tray and set it on a riser at the bottom of an insulated cooler.

The long answer? Palmer wrote it himself below.


How to easily make perfect clear ice cubes, repeatably, from a tray

By Mike Palmer

Here’s an easy way to make clear ice cubes in your freezer using an ice cube tray, obviating the need to carve individual cubes out of a block of clear ice.

What you'll need: A small "Igloo" type cooler that fits in your freezer. A flexible, silicon ice cube tray with holes punched through the bottom. Something to make those holes, perhaps along with a piece of soft wood. A paper or Styrofoam cup. A rock. And of course, a freezer.

If TLTR, skip down to “How to do it.”


I became obsessed with trying to make clear ice cubes at home (and later, clear ice spheres) after watching David Rees' How to Make an Ice Cube in his "Going Deep" series on the National Geographic TV channel. How hard could it be? (Ha!)

BTW, after finally succeeding at making clear ice cubes, I found that there are tangible benefits from using them in your drinks. (Even if drinking water.) In addition to lasting longer (because they’re pure ice and not ice/air), clear ice cubes also taste better because they’re pure ice. That is, they don’t have yucky tasting freezer air in them. Also, women like the aesthetics of clear ice and they notice—and like—the sound clear ice cubes make clinking in a glass. (Sure to please.) Freezer cubes don’t clink. They clunk. And now you’ll notice every time you hear a sound track in a movie or on TV.

In Theory

In David Rees’ TV program, he showed how clear ice forms in nature when a body of water slowly freezes from the top down with a substantial mass of warmer, unfrozen water underneath. (Alcademics readers knew this back in 2009. See

And he suggested on his website that you could make a block of clear ice by kinda imitating nature, by letting a pan of water cool in a freezer. Although that’s not really imitating nature since the water in the tray freezes from the outside in on all sides.

So freezing a pan of water does not make a clear block of ice, as even Mr. Rees concedes. Air still gets trapped in the middle, so you have to cut the “outer edges of your ice block” to harvest clear ice.


And even if the block were totally clear, you would still have to carve away at it to make individual cubes. And then you’d have to smooth six edges of each cube. that’s very time consuming, not to mention dangerous, working with sharp edged instruments on slippery ice.

Rees’ suggestion for making clear ice works better if you use the slow freezing method, using a dorm style fridge set to 30 degrees, as Mr. Kevin Liu has suggested.

[note: the above link redirects to spam, try the Internet Archive version instead:]

But you still get air in your cubes that way. Here’s a photo of an ice cube made that way. Air is still trapped in the last part of the cube to freeze. In this experiment, with only the center trays filled, that’s the bottom middle.

(I hypothesize that air is sucked into the cube when the last bit of water expands as it freezes. It lifts the cube up and pulls a vacuum underneath. Interestingly, air gets trapped in different places in the different cubes depending on whether a cube has a sister cube (or cubes) next to it. You can tell what position in the tray the cube had come from based on the distribution of air trapped inside it. In one experiment, I filled the tray in a checkerboard fashion, so that no cube shared a side with another cube. Only their centers were cloudy.)

In Practice

What to do? I tried the slow freezing method, which was supposed to do the trick. But it didn’t.

Fortunately we have the Internet. After searching around on youtube and seeing people consistently recommend “directional freezing,” I found Camper English’s Index of Ice Experiments on Alcademics. Camper was years ahead of everyone else.

In Camper’s landmark experiment, he found that freezing water in molds in a cooler (that is, from the top down) gave clear ice.


Except that air got trapped in the very bottom of the ice.

Why was that? Although Camper’s technique imitated the directional freezing aspect of nature, there was at least one other aspect missing: the large thermal mass of liquid water underneath the ice.

So, building on Camper’s experiment, I placed a silicon ice cube tray on top of a large Styrofoam cup (as a stand) in a cooler. I filled the cup and I filled the cooler to the top of the ice cube tray. (That brought the water level halfway up the cooler. More on the optimal water level later.)

I put the cooler in a high-end prosumer freezer, which was set at 0 degrees F, the government’s recommendation for food safety. (I checked the temperature in two places simultaneously using two digital thermometers in different locations in the freezer.)

The result of this experiment was much better, but I still got a little bit of air trapped in the bottom of the cubes.

If you look at a close up Mr. Liu’s ice cube, his is not perfectly clear either.

I hypothesized that there was still one more aspect of nature missing from this experiment, and that was that, unlike a Japanese ice pond, the water in the ice cube tray could not interact with the unfrozen water below it. One needed holes in the bottom of the ice cube tray for trapped air to escape.

At the same time I came to this “Eureka!” moment, I found that a reader on the Alcademics blog had made the same observation. Furthermore, this idea was consistent with a suggestion on the blog of suspending a spherical ice ball mold upside down, above a pot of water, with the mold’s fill hole in the water to make a clear ice sphere. (Or similarly, the suggestion to put a spherical ice ball mold upside down (hole down) in a mug of water so that the water in the mold could interact with the unfrozen water below in the mug.)

So here’s what I did.

How to do it


I took an acid brush and used the tube end as an auger to bore holes in the bottom of a silicon ice cube tray. Any similar tube will work. A 9mm shell casing would probably work too. (A .38 caliber casing is too large but works well on spherical ice molds.)

I got my trays from Amazon, if you look closely at the photos of the ice cubes on the Amazon product page, you can see air trapped inside them.

I used a Dremel tool with a cone-shaped cutter to grind a knife edge on the end of the tube. I put the ice cube tray on a piece of soft wood (to act as backing for the work) and pushed the tube into the bottom of the tray while rotating the tube. (If you use the end of an acid brush, you need to rotate it CCW so that the tube doesn’t unwind and open. If you use a shell casing, you can hammer on the casing (on the soft wood) to punch your holes through.)

Take a Styrofoam cup and cut it down to about an inch as a stand for the ice cube tray. (I started with a full size Styrofoam cup, but you’re just wasting water if you fill the cooler half full of water. And by using a one inch cup, the tray is lower in the cooler, resulting in better directional freezing.)

Put’s a rock in the cup to weigh down the cup. Else, the stand floats off the bottom of the cooler.


Put the tray in the cup that had the rock in it and filled the cooler with water to the top of the tray.

You might want to jiggle the tray a bit after you put it in the freezer, to knock air bubbles out that might be trapped underneath the tray.

Wait about 18 hours. Timing is somewhat critical. You want to catch the freezing process just as ice has formed below the tray. If you wait too long, all the water below the tray will have frozen and you might get air in your cubes. Remember, in the Japanese pond, there is always unfrozen water below the clear ice.


After the water under the tray has frozen, take the cooler out of the freezer, place it upside down in a sink and set a timer for about 10 to 20 minutes in case you forget to baby sit it. In about 10 or 20 minutes, you will hear a thunk as the ice block releases from inside the cooler. If you timed it right, a lot of water will eventually drain out of the cooler. (A cooler with the drain plug might be interesting. We could use compressed air to blow the ice block out.)

You will be greeted by a strange formation of ice crystals and a cavern of sorts from where unfrozen water drained. (Apparently the cooler isn’t as well insulated on its bottom as we would like, since ice forms there. Since cold air descends, shouldn’t coolers be twice as thick on their bottoms as they are on their sides? Maybe we should put a cooler in a cooler?)


Break the clumped ice away from the ice cube tray, push the cubes out of the tray from behind, and viola! Perfect, clear ice cubes.

Allow your cubes to temper for one minute before using them in your drinks to avoid cracking. 

Now you’re ready to do it all over again and make another batch!




High Strangeness

When I first tried these experiments, I started with the minimal amount of water, not knowing what the right amount of water underneath the tray should be.

I made a spacer that was only a half inch high.

When I did that, something strange happened. I got an extrusion of cloudy ice during the freezing process!


It has always been a corner cube that extrudes in all my runs. All the other cubes are clear. That’s gotta be telling me something.

So then I went the other way, putting more water in the mix by using a plastic beverage cup as a stand for the ice cube tray. I let the water freeze almost completely and didn’t get any extrusions. Let’s hear your thoughts about what’s causing the extrusions.


You can see how the water has frozen at the bottom of the cooler, again demonstrating that the bottom of the cooler is not as insulated as it should be. In any event, there’s enough clear ice below the tray to give us clear ice cubes. (Notice some air bubbles just below the tray.)

Ever being an engineer, I wondered what the “optimal” (minimal) amount of water was for this process. I kept shortening the cup until I got to one inch. Since I knew a half inch was too little, it seems that one inch is optimal. One inch seems to be about the same distance in the photo above too.

I can now make perfect clear ice cubes repeatably in less than 24 hours without a lot of waste.


Thanks Mike! 

Read about all the ice experients on Alcademics by following this link. 



Feed You can follow this conversation by subscribing to the comment feed for this post.


most original...

James Carroll

If you take a scissors to your ice tray, you can trim the trays so that they precisely fit the size of the cooler you are using. Minimizing waste and increasing the use-able cubes.

I can now get 24 (instead of 15) perfect cubes or 12 king cubes from each batch.

Like so:

Camper English

Thanks - I had not seen your post before, and neither did Mike, I'd imagine. Cheers!

John Donne

I just finished the insulated mug technique, but this post is making me go out and buy a cooler.

Camper English

We just need to convince Igloo to make coolers that are only a few inches tall. Or I could buy a big saw...

Doug Miller

Great post


Yep, the Igloo cooler method with holes in the bottom of the trays is what I detailed at my site It gets the ice almost-completely clear... plus there's usually some clear ice left over surrounding the trays I can carve small rocks out of, as well.


Reid Smith

This is similar to the way single crystal turbine airfoils for jet engines are cast. Interesting the way high tech aerospace manufacturing and mixology are related.


Does this work with distilled water?

Camper English

Distilled water is optional. You don't need it and it doesn't make clearer ice. Only use it if your tap water is gross.

Camper English

Wow, that's cool!

Mike P.

It didn't occurred to me that our coolers aren't insulated well enough on their bottoms until I wrote this article. So afterward, I put my cooler in a cheap aluminum foil-type cooking tray. Doing that helps insulate the bottom of the cooler a little bit. (I expect that cutting the tray and custom fitting it up tight to the cooler bottom (perhaps with HVAC metal foil tape) will perform even better.) I've only done one run using the tray, but I was able to use a half inch spacer again with less water and still get clear ice cubes.

Next project - how to drill a drain hole in the side of the cooler that won't leak when closed. (And that won't leak into the inside walls of the cooler.)


Mike P., instead of customizing and possibly compromising the cooler, could you insert a rod down the inside of the cooler? When ready to drain the cooler, pull the rod and insert some silicone tubing and siphon out the trapped water.


Here's what I'm going to try next, which should be the ultimate rig for doing this.

Take a silicone tray and punch holes in the individual molds.

Put the tray in a rectangular tupperware food container roughly 6x8x4 inches. Then put the tray/container in a styrofoam shipping box.

6x8x4.5 interior dimension (with 1.5 of surrounding insulation) seems to be a standard size insulated box available online from any distributor of boxes and shipping materials.

For extra bottom insulation (if you need it and have space), place the foam box on top of the lid (which you otherwise woould not need/use).

Should be a bit more space efficient than a hard sided cooler. Also, since the food container is clear, this would also have the advantage of letting you easily pull the container out of foam insulation to check on the progress of the freezing.


Maniac :-)

Camper English

Sounds good - keep me posted. I have on my to-do list to try a very similar thing but using that spray foam insulation where you're looking at the insulated box.



Wouldn't the tupperware tray prevent the water from interacting with the water underneath? I've been using the cooler/stand/pierced molds thing for a while now and it works wonders. 2-inch cubes are quite nice this way. By the way, in my setup it's the middle 2 cubes that rise up if I leave the whole thing in a bit longer than required.


Tupperware container is 4-5 inches deep and filled with water. Pierced ice cube tray for 2 inch cubes occupies just the top 2 inches. Works really well. I do get the same curious single ice column in one corner.

Basically the same set up as pictured above. Except that the water is contained in a clear plastic uninsulated container that slides easily in and out of a styrofoam insulated box.

Rob Albright

I theorycrafted this concept a while after your original post about using an Igloo cooler and the multicolor plastic pieces that you used in one of the pics above.

I'm very pleased that it worked out without the need for cutting any/(much) ice.

I decided that I couldn't scale it into my available freezer space at the time and had to abandon the idea.

For the home bar enthusiasts, I would suggest the 'Tovolo King Cube Ice Tray'(search Amazon; can't link from my phone. Sorry.) Easy mod; impress your house guests, enjoy a whiskey rocks or something along those lines.


Rob Albright

@Camper Avoid the corners. A single tray, centered, elevated, ought to work. Directional cooling is the correct answer.

Otherwise, make friends with someone with a Clinebell.

I would presume the extrusion in the corner is a result of expansion and the vacuum forming directly below the cube, which you mention in the article. As the cube freezes, it lifts up a bit, and water from beneath the tray (as well as air from above it) get sucked in through the hole. That water then freezes, and the resulting expansion and vacuum again lift the cube up. The process would repeat until all the water is gone, the cube runs out of space, or the forces of gravity, buoyancy, and vacuum balance out.

I'm not sure why it only happens in the corner, although in that spot, perhaps there is less room for circulation of air and heat. Maybe I'll look at this in greater detail some time soon. If I do, I'll let you know what I come up with.


To elaborate, the thing to notice about the cube in the corner is that it has the least "degrees of freedom" in terms of where nearby water can go. If we consider a cube in the center or on the side, let's say it can flow in four directions: north, east, south, west, or whatever. On the side, it can only flow east, south, or west. In the corner, even more is lost, and it can only flow south or west. However, increase the water level, you've now added another dimension: down. In this case, you no longer see the extrusion, correct? So perhaps when there are less than three "degrees of freedom" (or lack of a certain combination of them?), it results in the vacuum pulling and growth of the cube.

Camper English

Very interesting, thank you!

Mike Forstadt

Great article! One thing: a 9mm case is .380 inches at the neck. A .38 special case is .379 inches at the neck. They are virtually the same size, but the 9mm case is actually 1/1000 of an inch larger (even though the bullet is 2/1000 of an inch smaller) than the .38. Would there really be a difference for the purposes of this application?

Camper English

Oh I doubt that makes much difference- it's just a matter of whether the hole is so big that your ice cube tray is mostly hole and not enough tray :)


Yes, fewer degrees of freedom, but then why not all four corners instead of just one? I think it has more to do with the cooler was not perfectly level when placed in the freezer. I also notice in the extruded ice photo that the ice cube tray is closest to that corner of the cooler. Coincidence or causation?


I jury-rigged this by putting the king tray in a tightly fitting glass tupperware tray, and then putting that tray into a larger tray (+1" on each side) and stuffing the space with dish towels.

I got the extrusion too, and it looks like that corner was the last area to freeze. Basically, that last bit wants to expand, but the whole top is frozen, so it just pushes out that last corner.

vincent thuaud

One this is done, would it alter the quality of the ice if I would empty the silicone tray in a plastic bag and stack the ice cubes multiple times in the freezer?

I need to make clear Ice in a commercial/bar setup.

Let me know your thoughts!

Camper English

Ice cubes don't get more or less cloudy once they're frozen. So yes, this would work, though they still freeze slowly and one tray at a time. Note that ice does pick up aromas and evaporates in the freezer so be sure to make it a sealed bag/container you store it in.


Good pun



Im inquiring your expertise on building an ice igloo. I've collected about 500 milk cartons (2L) and over 250 (1L) milk cartons. Id like to know best to freeze the water so that the clarity of the ice is close to 100 %. Thank you in advance.

Camper English

Hi - The only way it will probably work is to put those cartons in an insulated container so that they freeze only from the top-down. That would get it at least mostly clear. Take a look at all the experiments here for why

Ayy Lmao

Just use a cheap styrofoam beer cooler

Jim Showalter

I get that same extrusion problem using Camper's technique. Haven't been able to tie it to any particular thing, and it happens sometimes and not other times. Has anyone come up with a theory?

Camper English

My guess is that it's an extreme version of Ice Spikes:

With the interesting paragraph here:

"Small ice spikes can be formed artificially on ice cubes produced in domestic refrigerators using distilled water in plastic ice cube trays. The formation of the spike is similar to that of the naturally occurring spike in that the expansion of interior water and the reduction of volume in the interior of the cube increase the pressure on the water pushing it upward through the hole. The growth of the tube ceases when the drop at the top of the tube freezes entirely, which is substantially before the rest of the water in the cube is frozen.[4] This method produces small spikes which are usually round or triangular in cross section with sharp tips. Experiments using this method have been carried out in laboratory settings but it has been found that spikes are less likely to form in ice cubes made from non-distilled water as impurities in the water inhibit spike formation.[nb 1][10] This poses the question of how naturally occurring ice spikes form in tapwater or rainwater and Libbrecht and Lui have suggested that, in the case of the small spikes grown in a refrigerator, impurities will become increasingly concentrated in the small unfrozen droplet at the top of the tube reducing the freezing rate and so the growth of the tube. However, they believe that on the rare occasions when exceptionally large spikes grow in natural, outdoor ice formations, some other mechanism must remove the impurities that build up at the top of the growing tube. Either impurities may be forced into pockets that freeze more slowly, or perhaps a convective flow, which would be insignificant in the smaller, artificially grown spikes, replaces the water at the top of the tube with fresh water from below.[3]"

If CalTech hasn't figured it out yet, maybe we can :)

AJ Peacock

It seems like the extrusion is caused by the pressure created by the water below the tray as it freezes. Once the surface of the entire water freezes, the pressure created below the surface (as the ice freezes) causes the remaining liquid water to take the path of least resistance (up through a hole, pushing up the cube).

I suspect that the corner ice cube is easier to move upward, since the silicone tray can stretch easier on a corner cube (less support by neighboring cubes).


That's what I was thinking. I don't have enough room to put a whole igloo cooler in my freezer compartment, so I thought of using the foil-backed styrofoam insulation that I've used to insulate roof spaces and gluing together a make-shift "cooler" just the right sizefor one tray of cubes, plus it's outer tray to hold the water bath. this might fit into my freezer compartment and allow me to do one tray at a time.
But I have to say, EXCELLENT web post! Lots of good ideas!

Adrian St John

I have been using this method with Zoku Ice Sphere Molds and have been experiencing the extrusion as well. I am only making two spheres at a time and the extrusion manifests in only one of the molds. I get one perfect sphere and one Saturn-shaped sphere (as it extrudes the mold opens up and a belt forms). I am in the camp that it is because this sphere is freezing last and thus is being affected by the unfrozen water beneath it. I'm going to try putting some sort of mold alongside my spheres such that it is deeper in the water and will therefore, hopefully, freeze last and be the outlet for extrusion.

Jeffrey Sirninger

I bet the extrusion occurs due to the use of flexible silicon trays secondary to differential expansion/compression between cells. For instance I highly doubt this would occur if you used a metal or hard plastic tray. Basically as the partially frozen ice starts to expand some cells will freeze slightly slower and be subject to more compression by adjacent cells squeezing it and eventually extruding it.

By any chance is it a positional phenomenon, either in relation to cooler position and/or to certain cells within specific silicon trays?

Camper English

In my experience it tends to happen roughly in the same place each time - away from the fan/cooling element, on the side towards the freezer wall, in one of the last two cube slots. Thus, likely one of the last places to freeze. Which supports the running theory of the tray squeezing everything into the last space and lifting it up.

Jeffrey Sirninger

...all of this occurring with the perimeter on the normally flexible tray being rigidly constrained by surrounding hardened ice, so the expansion leads to compression that must be released somehow.

Micah D.

When you put your plastic homemade ice molds(the colored ones from the Container Store) into the cooler are you using anything to elevate them or just placing them in direct contact of the bottom of the cooler?

Camper English

I'm just putting them on the bottom of the cooler, so if you let it freeze all the way they are cloudy on the bottoms. Note that these make cool spears, but
1. They're not graded food-safe, so I wouldn't serve them to customers in a bar situation (takes too long to make them anyway).
2. They're hard plastic so they crack as the ice expands and will eventually get destroyed.

A more ideal "tray" would be a tall silicone tray like the one on this post, with holes poked in the bottom.

Reese Aptor

Hello, Mike,

I did your experiment and it worked perfectly; I needed to get clear ice cubes to make the point about fog and clouds reflecting light back to us, whereas frozen water without air bubbles allow light to pass straight through (I reference your work in my book that just came out: "Turning Fog into Beer".


Reese Aptor

Joe OBrien

Thanks for all the info & testing info!! I am very intrigued by all of this! I love the "science" of cooking -- Kenji Lopez-Alt is my new Alton Brown :). So, making clear ice cubes based on the science of directional freezing caught my eye. The igloo method is definitely the way to go, but my freezer is VERY small and space was limited. I happened to have a small rectangle Tupperware container on hand (about 9x6x6). I also had some 1/2 in craft styrofoam. So I built my own small "igloo" cooler with the styrofoam exactly the size to fit the Tupperware in, and covered it all in aluminum duct tape. From a directional freezing standpoint, the results have been great! It's very compact! I have been using silicone ice trays (1" and 2") with small holes in each and using a few small plastic food containers to lift the trays up. I let it freeze for about 12 hours or so (I haven't perfected the timing yet). I am still getting little streaks of air, but the cubes are virtually clear (but I think it's user error). I am still testing it all out -- even trying to NOT use a tray, but just let the top 2-3 inches freeze and make my own "artisanal / hand-crafted" ice cubes!

Camper English

Thanks Joe, glad the site/post is of use!

Christopher Werby

Just found your interesting experiments. I have also been trying to get clear ice. I've found success with a little insulated box containing an ice cube tray that just keeps the top half of a frozen column of ice -- almost exactly your approach does. Maybe the manufacturers read this post! I've been using the device for about 3 months with good results, although I'm always afraid that the silicon is going to rip -- you need to do a lot of manipulating to get the ice cubes out of the mold. And 18 hours is exactly the right amount of time.

Camper English

Yep I'm pretty sure that allll the clear ice tray manufacturers learned it here or from someone who learned it here. December will be my 10 year anniversary of figuring out the basic method! (Happy iceversary to me.) Anyway if your device does rip, check out the Clearly Frozen tray -


hey, very cool post and nice job experimenting. however, i have a real bone to pick with your hypothesis about air getting sucked into a cube through a vacuum. that’s really not how it works at all. all liquids have gas dissolved in them. the hotter the liquid, the less gas it can hold, which is why boiling water is another trick for clear ice. colder liquids hold gas very well, right up until they freeze into a solid, which is why we see distinct air bubbles. and the reason that all the air bubbles clump up into one place is that ice is a highly crystalline structure, and it is much more stable for it to grow in its crystal structure without small air molecules disrupting that pattern, so the air is pushed to parts that have not yet frozen. then, the air is boxed into the ice because all sides have frozen and the air can not escape. this is why clinebells make clear ice: by only freezing one side, dissolved air escapes through the unfrozen water.

another, possibly easier way of utilizing directional cooling for clear ice would be to take a normal ice tray, fill it with water, and insulate the top by covering it in a large piece of foam. because the cold freezer air cannot circulate around the top surface of the water, the bottom and sides freeze first, pushing the dissolved air upward. i think that as long as the foam wasn’t an incredibly tight seal against the water, the dissolved gas could escape before the top finally freezes. it would involve a lot less work than this method. hopefully someone can try it out!

Jackson Dewey

Great post. One dumb question. I want to do this technique with a silicone mould that I made and am going to cast some ice rather than beeswax. I was planning on inserting a very thin cocktail straw through a hole in the bottom of the mould and am wondering it the water flow from the cooler to the tray (in my case my mould) needs to be a minimum size to make everything function properly? Any suggestions would be greatly appreciate as I would like to keep the size of the hole as small as possible and only penetrate the mould once rather than testing. Best, Jackson

Camper English

I'm not sure that I understand the system you're talking about, but in Dave's model and the one I did to replicate this the holes I poked in the silicone were quite small; probaby cocktail straw size, and it worked fine.
Pics here:

henry c stage

I wouldn't take the time to make the acid brush punch. You can get open punches like that from Harbor Freight for next to nothing. Good for leather, rubber, whatever. Cheers!


Hi there, I just wanted to stop by with an idea and a question. Before I get to that though, I want to thank you for your website - it's where I learned how to make clear ice, and refine it too.

My current setup is a cooler, two 2" x 6 cube molds (ie. it churns out twelve 2" cubes each time) that I rest on 9oz solo cups (they're ~2.5" tall.) This setup works nearly perfectly, save for one issue; the pressured air-spikes. That's the biggest issue with this method is the way it seals the top off and then pressurizes the rest of the system, and it's one of the biggest reasons that things can go wrong. Sometimes I'll get one cube that slips a bit and extrudes, which releases some of the pressure and saves the system a bit, however that's not always the case, and I'll often get the air spikes.

To be clear, <--- this is what I am talking about when I'm saying "air spikes." This is the very exaggerated version you get when you let it freeze all the way, but even leaving 2.5" of water in the cooler with my method gets a bit of this action sometimes.

The solution is pretty clearly a proper pressure relief. Unfortunately, as you found out ( this isn't really an easy thing to accomplish. Having said that, I do have an idea. If there were a way to create an insulated straw of sorts AAAAAANNNND place it in the cooler in the area that freezes last (probably furthest from the cooling element / fan in the freezer AAAAAND the insulated straw were weighed down so it couldn't be forced upwards and out, AAAAAND the inside of the straw was very low friction, then I imagine it may work as an extrusion / air release a bit more reliably and one might even be able to go unplug it about halfway through to encourage further pressure release.

Follow me here for a second as I go through my thought process on the features of this pressure release valve.

> Feature 1: Placement where it freezes last -- we've established that the extrusion occurs where things freeze last. My guess is it occurs this way due to the "squeezing" (mentioned in the comments) AND because the other cubes already have a bit more ice, and therefore have greater friction (resistance) to overcome before moving a bit, so the last to freeze is the most likely to slip as the pressure builds. Once one has slipped, it's going to be a lot easier to continue moving it, compared to the others that are getting progressively more difficult to move.

> Feature 2: Insulation -- again, we're trying to get this sucker to freeze *last.* While the system is obviously going to circulate water, if we can isolate and insulate juuuust enough to be slightly higher than the rest of the system, then it's going to be more likely to freeze last. Honestly, whether this bit actually makes a difference is questionable, and since I don't actually have a good way to insulate, I'm not sure this part is really realistic (or necessary.)

> Feature 3: Weighted / Anchored -- this part is definitely going to be important. The straw used in the aforementioned pressure release experiment was being pushed up and out. An anchoring or weighting method needs to be used to keep this thing in place without a doubt.

> Feature 4: Low Friction -- this doesn't need much explanation, really at this point. We want the ice that forms in here to slip more easily than any other surface in the system. I don't really love the sound of Teflon, and I'm guessing there are no Teflon coated straws out there. Other than that, I don't really know what other low energy / low friction type surfaces could be used with a tube like this, or would be readily available.

I unfortunately am here only with ideas, and not solutions. If any of you have thoughts on how to make this pressure relief setup a reality, I'd love to hear it!


Oh, by the way, there is probably a way to resolve any of this fussing. The air spikes and pressure issues would pretty much be alleviated if one were to do this in a soft insulated container that was able to expand with the ice instead of rigidly resisting it, and forcing the pressure to build. I haven't really bothered looking too far into this option because I don't think the insulation in soft-sided coolers is sufficient, and it may end up breaking down those often-flimsy liners even before they'd normally break down.

Still, if someone has a bright idea about how to move away from the hard coolers to a less rigid system easily and inexpensively, please let me know.


Just commenting with my newly created TypePad account so I get notifications if someone actually responds.

Camper English

Thanks kda0000 - I've been having 'problems' with what I've been calling the Mystery Pillar - the last cube that pops up from this post:

But I haven't done any brainstorming on it. So I'll do more of that here.
- Put a ziplock with some air in it underneath the trays and maybe it will pop open with the pressure of freezing... but then where will that trapped air go?
- something else collapsible, or squishy like foam beneath the tray
- have a sacrificial cube of some sort... I can't think of how this would work
- Have the straw be something soft and insulated like a rolled up piece of foam that can squish...

Those are first thoughts, maybe I can give a few of these a try but please do let me know if you attempt and find anything that works!


Well, I guess my hopes of being notified by TypePad didn't work out, though maybe that's because an email notification got lost in the shuffle of the hundreds of spam emails lol. Anyway, I think your ideas are probably better and more likely to work than my thoughts about an pressure release system. Probably the most feasible way to accomplish such a pressure release would be to design an ice mold that specifically focuses on sacrificing a cube. Maybe one of the cube slots might have a very low surface energy coating that would allow it to slip.... although I think silicone is extremely low as it is, so I'm not really sure what could be used, and that's probably way outside the scope of this discussion anyway.

I like the thought about creating an air pocket below the trays which can then be compressed. Unlike water, air can actually be compressed, and I have to imagine that it might alleviate at least some of the issue here. With that in mind, I think I may try an inflated ziplock bag weighted down with something heavy enough to inside to pin it to the bottom of the container so that it doesn't float up and disrupt the surface (maybe some whiskey stones or something.) I think I'll need to be a bit careful to position the opening of the bag to point up and sit directly between the two cube molds, because if the pressure does cause it to burst open, I don't want that air escaping up into the cube molds and creating air pockets in the cubes.

Really though, the more I think about it, the more I am just considering sacrificing my cooler and doing this with it --> I have a feeling that this method could be coupled with some silicone trays to make completely clear cubes without any carving. I think with the ones I already have, I could just flip them upside down and weight them down to sit at the bottom of the cooler, being careful not to block the venting holes. I'm guessing that the lack of direct circulating air contact with the water itself, plus the slightly larger volume of water required to avoid pump breakage might mean 5 or 6 hours of added time, but that's no problem. I guess my primary concern would be how long the pump would last and concerns about how to get the cord out of the freezer without having to worry about the seal.

Helen Ayres

FYI, the sciencefare link is now redirecting spuriously to apple's iTunes website.

I suspect some jiggery-pokery, but thought you should know so you can removed the link after checking it yourself of course!

Camper English

@Helen thank you! I'll see if I can fix the link.

Verify your Comment

Previewing your Comment

This is only a preview. Your comment has not yet been posted.

Your comment could not be posted. Error type:
Your comment has been posted. Post another comment

The letters and numbers you entered did not match the image. Please try again.

As a final step before posting your comment, enter the letters and numbers you see in the image below. This prevents automated programs from posting comments.

Having trouble reading this image? View an alternate.


Post a comment

Your Information

(Name and email address are required. Email address will not be displayed with the comment.)