"Dr. Osborne experimented on the issue back at his workplace and confirmed Mpemba's finding. They published the results together in 1968.

...

The effect of heating on dissolved gases; however, this was accounted for in the original article by using boiled water."

-wiki

Has anyone here tried to test this "claim" ? I mean give me a break i tried this evening to freeze faucet water and it took 160-164 min (obviously i couldnt be opening freezer all the time so i am guessing a bit earlier than i opened it last based on the time of the prior test and the condition of the ice that had started to form from the perimeter of the can first to full disk eventually to build 2 mm thick ice wall at the top (water below ) (so lets define this as the reference). I use a metal cylindrical can of 500ml volume probably filled to about 450 ml. Then used boiled water when down to around 60C still steaming and put it in the same location in the freezer same can same volume and got the same type of thickness fully formed solid top surface of 2mm at about 170-174min ie about 10 min later. Guess what, i bet that extra 10 min is the time it takes for 60 to drop to 15-20 that faucet water is, lol.


So there you have it do it yourselves, my setup is ridiculously crude but faithfully done with as good conditions as possible and i got the hot water to freeze slightly later.

So yes i call BS on this claim! LOL

Do it and prove me wrong. Is this a bs hoax now or did i get bad results. But how on earth can my results be bad when it took almost 3 hours to freeze in both cases ie a very slow process that within any experimental error can most definitely be correct within 5% since 3 hours is enough time for freezer to go over all possible cycles it does during the hour fairly similarly between the 2 cases.

That said maybe they have in mind some super fast freezers not the ones people have at home. But whatever, i am waiting your own results guys.

Lets define freezing at these 3 stages;
(takes at least 2 hours for me to get to part 1)
1) Early on water gets very cold and starts feeling heavy (as if it has higher viscosity)
2) A while later ice forms in the perimeter first in small segments
3) Several minutes later ice of depth 1-2 mm forms that covers the entire top exposed water surface of the can.


Go ahead and do it as well and tell what you get.

I am not surprised at your result, but I am not ready to call the original claim BS either. I still think the result of a specific test is going to depend on the geometry and moc of the container, the purity and history of both water samples, the temperature of the freezer and the starting temperature of the samples. I would bet that one could construct experiments that turn out both ways.

Let me give one example. Take the case of a hot sample at 100C and cool sample at 25C. Initially, virtually all of the heat transfer will be due to evaporative cooling at the surface, particularly for the hot sample. By the time it cools to 25C (which will happen quickly given the large driving force for cooling), the hot sample will have lost ~15% of its mass to evaporation. Therefore, it should take about 15% less time for it to get from 25C to frozen (all other things being equal), then the cool sample. If the freezer is not real cold, then the 15% reduction in the slow process of freezing could possibly be more important than the relatively fast process of cooling from 100C to 25C. MOC of the container is important because the more insulating the container is, the slower the freeze process and the more important the evaporative mass reduction is.

Try your experiment in a glass or plastic container. You might get a different result. Also, fully frozen versus a layer of ice is important. I am not sure how the original effect was measured.

It seems like if you want to run this with cheap lab equipment, you should boil a big sample of water, then pour it into a volumetric flask to measure and then immediately into a much larger florence flask and stoppered (a mostly empty florence flask so freezing is less likely to blow it out because the ice will have room to rise as lower water freezes.. if it's actually safe to freeze water in an erlenmeyer flask then just do that instead, or if you're only going to the start of freezing it won't matter). Once you have two of those, then let one cool down however while keeping the other one hot. Then wipe them both down, put them in a freezer with a lot of desiccant (so there are no exterior frost effects) and see what happens. You can also run a thermometer through the stopper to measure temperature along the way.

That should correct for any meaningful difference in water volume (it'll reequilibrate with the vapor in the flask pretty fast as it cools as opposed to shedding a huge amount of mass to the outside world) as well as standardizing anything dissolved in the water to begin with, or after the fact while they're different temperatures. Alternating which florence flask gets the hot/cold sample should expose a container differences if any.

Well i can honestly say that the 2 water samples were both almost at the same height in the can indicating pretty close volume to say 2% really and evaporation therefore is very little coming down from 60. Even from 100 i dont expect it to lose more than 3-5%. While boiling of course its fast.

Of course i will try many variations with whatever crude equipment i have at home but i was expecting based on claims to see something spectacular and when i didnt and not only that but the hot one was late by 10 min or so it felt as a joke really.

But i agree we dont know how they defined freezing and maybe the total volume freeze is a different outcome. However i expect once the first ice forms on top the rest to proceed similarly between the 2. Why? Well because 3 hours is pretty damn stupid long period of time for the water to basically become nearly identical state in the 2 samples. The only difference left therefore would be no dissolved gases on boiled and maybe slightly different concentration on minerals (tiny change).

Maybe the effect is very different in rapid cooling cases. But what on earth rapid cooling processes would Aristotle have from all mentioned back then 2360 years ago? Only natural ones from cold weather (gets rarely -10 C where he lived) which he would then need what? to throw hot water in the snow or a frozen lake and see what happens???

Maybe if the cooling happens ultra fast by forcing water to be in significant contact with surrounding mass at whatever -15 C the freezer is, the effect is not just faster but a lot faster. I will try plastic bags of water placed around meat for maximum contact and see for one how faster the result is. 3h was pretty damn ridiculous!

Check your freezer temperature. Taking almost 3 hours to get a 2mm thick layer of ice sounds like it's taking way too long. It would take forever to make a tray of ice at that rate. (Or maybe I'm just misreading your run-on sentence.)

I think household freezers should be running at close to 0 degrees F.

No my freezer works ok although old. It is about -15c and it has a lot of meat and frozen vegetables ice cream etc in it so its not empty which should help i think. But when you place a can and its not in contact from all sides but only with the air of the container and the bottom part of its surface it takes a long time.

Also a tray of ice probably happens faster because the contact area for the water you use is a lot bigger in relation to the volume. Its like 5a^2 vs a^3 and for the cylinder its 2*pi*R*H +pi*R^2 vs pI*R^2*H. So if a for the ice tray is 5-6 times smaller than the cylinder's R or H you get the picture. Also the air inside the freezer must quickly arrive at the overall temperature of the system when opened and closed due to small overall mass exchanged. So maybe one can use all surface (eg above equations) although i expect the metal contact to be doing much better job in cooling than the air molecules above.

Something just doesn't seem right to me... 3 hours to get 2mm of ice on top of a half liter of water just seems really slow, and this has nothing to do with the Mpemba question.

Try it, put in your freezer similar volume in a container and let it rest and report how long it takes. I know my food is pretty cold because it takes forever to thaw when i take chicken or steaks out which wouldn't be the case if the temperature was very close to say -2 or -3. It it is indeed pretty low as i said around -15. But as i said when you put food or ice trays you let the mass rest having very small height (in packets) so it has much more available surface area for its mass than similar mass water in a cylinder. I mean if as i said i try to place a plastic storage bag filled with 100ml water resting flat on the surface of the freezer it will obviously take 1h or so maybe less. In fact i have frozen vinegar as part of another test within minutes in a small r=2cm disk cup-height 0.5cm (way much less volume) . Of course not having the exact same surface in both runs makes the freezer bag example a problem to test.

In any case none of these bs in my opinion references in wikipedia describe what they did and what they got as results. So i call these references bs not because they are lies necessarily but because as presented they do nothing to illuminate the reader on what the hell they measured. Wouldnt you expect some numbers in an article like that?

For a phenomenon like that in the media recently you would expect an article that has some exact data in it and yet where is it??? Give me some numbers minutes , hours, how it was done, what was defined as frozen to stop time etc.Where is all this in the wikipedia entry???

I will do a lot more tests now that i am curious but you also try your own and report so that we get a better picture.

Maybe i must try a lot less water next in a much different shape container (eg a small disk plate like the vinegar thing above to see if then the fact it happens a lot faster defines the difference better.

Still curious how on earth Aristotle would have known any of it using what cooling method? lakes? snow?

Ok lets try to read this article to see if it has more illuminating details about what the experiments are like and some theory too.

http://arxiv.org/pdf/0704.1381v1.pdf

"Through the ages, the seasonal harvesting of snow and ice was a regular practice of most of the ancient cultures: Chinese, Greeks, Romans, Persians. Ice and snow were stored in caves or dugouts lined with straw or other insulating materials. "

-wiki

As I said earlier, household freezers should be running at about 0 F = -18C, so you're actually little warm. I don't think it's warm enough to make a huge difference in freezing times, but your freezer is not running "pretty low."

Also, meats freeze solid at temperatures lower than -2 or -3 C because there are lots of dissolved solids in the water of meat products. I think turkeys freeze solid at around 25 F = -4 C or so.

So i guess ice harvesting even in Greece could work probably north in the real historical Macedonia that Aristotle spent some time of his life and where winters bring snow that lasts many months like it does in North America (but of course never in coastal/central California etc only in mountains).

This looks impressive too for the old timers;

http://en.wikipedia.org/wiki/Icehouse_%28building%29


Which of course makes it interesting now to ask;

If i have a mountain north and can bring down to the valley snow and store say 90% of it before it melts in a properly insulated house how long will it last if the weather outside is typical 20c or 10C or even 5C if its say not summer. I mean i read the references but i still would like some real numbers of how it worked (since it was in use until modern times and they could therefore check it then)

But my argument is that if Aristotle and others were able to use even without describing to you how they personally did it without modern refrigeration the best they could have is some snow house at -2-3C? If they still could experience the effect it must be pretty easy to reproduce in a freezer today!

http://www.jphs.org/locales/2004/6/2...aica-pond.html

I wonder what kind of low temperature they could have achieved with the ice houses anyway. Some of the insulation used was pretty thick.

I think the first mistake is to think of the ice as being at close to 0C to begin with. It was probably much, much colder. This would make the ice house much colder than -3C. It's not that the ice was "maintained" at a constant temperature. It was very cold, and then over time it got less cold while still being colder than 0C.

Also, have you lived anywhere that it snows? Have you ever seen snow and ice sit in a shaded valley for a week without melting*, even though it's not that cold out?

* That is, without completely melting away.

My grandfather used to work on an ice camp in northern Wisconsin driving teams of oxen. They would harvest ice from frozen lakes to sell in the cities to use for refrigeration. People would have "ice boxes" which in those days were literally insulated boxes that held ice, and the ice man would come and deliver the ice every day.

Oh i am not disputing that meats freeze easy at -3-4 etc (most food is water anyway) , just that my meats take forever to thaw something that doesnt happen if they were just -3,-4,-5. So yes i am sure its <-15 or so and i will measure it next and find out for sure but -15, -18 or -24 that i have had before isnt the main issue, geometry and placement and size are!


Back from measuring mine where i did the freezing and came at -17 C or so, the -15 C was a guess. The point being that the equipment was capable for that low, its simply that in my first experiment the process was slow due to the size/geometry and neighboring material contact of the container that had the water. As i said a small plate or even a full spoon will freeze water much much faster than a bottle or can.