Cash prize offered for solving Mpemba effect

krd
What puzzles me is: the temperature of the 'hot water' will drop. On the way, it will pass the temperature that the 'cold water' started at. Unless there is some other factor that the experimenters haven't told us about, it is then the same stuff at the same temperature that the cold water was (it is now 'cold water') but later. How can it then (from that temperature) cool faster and overtake than the other lot of water?
Think of it in terms of two buckets - a warm bucket and a hot one - most of the heat is lost at the top of the bucket. In the warm bucket there's a warm trapped layer in the middle of the bucket - it's trapped by falling cold water. In the hot bucket the middle layer is more energetic, and breaks through to the top of the bucket where it cools quicker. The hot bucket even pushes down cold water with more force. The distribution of heat in the two buckets is completely different.

IF.......You had an stirrer going in both buckets......the cooler bucket would freeze first.

I heard about this competition a while back. I think they were looking for a fancy pants demonstration. Doing it with thermal cameras and a walk in freezer might be a thing - then quite a bit of experimentation to get the optimal temperatures for the demonstration. It will probably be done by some precious rich kid nerd (with lots of help) who will then go on TED Talks, and make us all want to vomit.

https://dl.dropbox.com/u/22026080/IMG_8810.JPG [Broken] was taken right before the touching test

#1 and #2 are precooled boiled tap and rain water |(that did not react upon touching)
#3 is plain, not precooled tap water (apparantly already partly frozen)
#4 and #5 are initially hot boiled tap and rain water (that flash froze upon touching a few seconds later).

Something trivial :P struck me there.
I wanted to make clearer pictures so I wiped the condensated water off the tubes and then...?

Last edited by a moderator:
krd
I wanted to make clearer pictures so I wiped the condensated water off the tubes and then...?
I think your tubes may be too small.

And before doing these, or any other, experiments, you're meant to have some kind of theory.

krd
The hypothesis about impurities like dissolved gasses, delaying crystal forming appeared reasonable.
But there should be less dissolved gasses in the hot sample than the cool one.

For crystal nucleation total absence of impurities is more of a problem. Once the crystals nucleate they get going. Beyond a certain point it doesn't really matter how much impurity is in the water.

The Mpemba effect is a conversation piece. It would be a real pain in the *** to actually replicate. Though I think I may have seen it as a child. I think it may be important in the way jelly/jello sets in the fridge.

And before doing these, or any other, experiments, you're meant to have some kind of theory.
So they drill in your head in high school science class on the scientific method. I don't think he is planning to submit his observations to any peer reviewed journal. Also, hypothesis != theory, even though that's not even necessary. It can be a helpful tool to guide your experiments, and if you want to prove something, you need one, but there's no scientific rule that says you need to have one to simply make observations. Even experimenting, by itself, is not the scientific method. You don't need to have a theory to make controlled observations. Many theories came after observations.

Do you think Maxwell's equations existed before Faraday?

Last edited:
krd
I don't think he is planning to submit his observations to any peer reviewed journal.
No, but he's trying to repeat an effect that has already been observed.

The whole Mpemba combination - and I think they have prizes for a whole bunch of similar problems. It's not just that should be able to repeat it, but you also need a theory to make it repeatable.

I think it might take quite a deal of messing around to find the ideal conditions.

Once an effect has been observed - then you need a theory - if the theory allows a reproducible result, then maybe you have something. But it has been reproduced. If I was going to do it, I'd try to track down any of the notes or people who'd been involved. Set up the conditions to repeat it. Look into the body of knowledge that might cover it - set up my thermal cameras and buckets in a walk in freezer.

But there should be less dissolved gasses in the hot sample than the cool one.
Sorry if I confused you, we completely agree there. The idea is that hotter water freezes quicker because of that effect.

But with confusions like these you may wonder what causes you to be on the wrong foot.

Oh and I submitted those observations I did in he second test, I'm just wondering if anybody sees what I see, with the key word 'condensation'

https://dl.dropbox.com/u/22026080/IMG_8810-1.jpg [Broken]

Last edited by a moderator:
krd
Oh and I submitted those observations I did in he second test, I'm just wondering if anybody sees what I see, with the key word 'condensation'
I don't think you're onto anything there. The condensation will be thermally neutral, and if anything may act as an insulator.

Your cool tubes, were cool enough to allow condensation, and not hot enough to stop it.

You know when you get wet, you feel cold?....That's because the thin layer of warm air is no longer over your skin. So, your body cools quicker. But that principle is not a factor in the Mpemba effect.

I think one of your problems in the reproduction of the Mpemba effect is your tubes are too small - too much surface area in relation to your volume.

russ_watters
Mentor
I did a test a few years ago with coffee cups. Results still negative.

The condensation will be thermally neutral, and if anything may act as an insulator.
I don't think so. It appears that the latent heat of condensation is about 2500 joule per gram that is added to the sample. But it is clearly visible that the cold start samples #1 and #2 had much more condensation going on than the warm samples and that would have to decrease the rate of the cooling, latent heat.

https://dl.dropbox.com/u/22026080/IMG_8810-1.jpg [Broken]

So I made a new entry suggesting that whenever you find the Mpemba effect, check for condensation. Maybe the effect is just latent heat of condensation.

edit added the pic again

Last edited by a moderator:
No, but he's trying to repeat an effect that has already been observed.

The whole Mpemba combination - and I think they have prizes for a whole bunch of similar problems. It's not just that should be able to repeat it, but you also need a theory to make it repeatable.
I'm really confused as to how a theory makes an observation any more repeatable. Anyway, I was impressed he went to the effort to try to reproduce the observations, with or without a theory.

I have done a sloppy version of what Andre did in the past, same as Watters, and nope.

I tried to do a better test, not as well done as Andre. Nope.

If anyone can actually get this to work chime in please. I have read the papers, but a simple procedure would be nice. I have never observed liquid CO2 either even though I know others have been able to reproduce the pressures, etc... needed. Is this gonna be as difficult for the layman?

* I love to watch the flash freezing. Freaks me out every time I see it. On the other hand, I also rather enjoy being shocked by a Van de Graff...

russ_watters
Mentor
The Usenet FAQ has a very good entry about the Mpemba effect:

http://math.ucr.edu/home/baez/physics/General/hot_water.html
From the article:
Hot water can in fact freeze faster than cold water for a wide range of experimental conditions. This phenomenon is extremely counterintuitive, and surprising even to most scientists, but it is in fact real....

It is still not known exactly why this happens...

Why hasn't modern science answered this seemingly simple question about cooling water? The main problem is that the time it takes water to freeze is highly sensitive to a number of details in the experimental setup, such as the shape and size of the container, the shape and size of the refrigeration unit, the gas and impurity content of the water, how the time of freezing is defined, and so on. Because of this sensitivity, while experiments have generally agreed that the Mpemba effect occurs, they disagree over the conditions under which it occurs, and thus about why it occurs.
My problem with this is that there is an implied repeat-ability problem here and if the experiment is highly dependent on initial conditions, but isn't repeatable, then you can't with confidence say that it is real and not, for example, a product of experimental error. See: Cold Fusion.

You know when you get wet, you feel cold?....That's because the thin layer of warm air is no longer over your skin. So, your body cools quicker. But that principle is not a factor in the Mpemba effect.
If that's the reason you feel cold when you get wet, how do you explain why it feels cold just after you step out of a warm shower? The water you're covered with is warmer than the air in the room, so why should you still feel cold when the "layer of warm air" is just replaced with warm water?

The real reason you feel cold when you're covered in water is because of the latent heat of evaporation of the water. As the water sits on your skin, a small amount of the water molecules gain a kinetic energy that is much higher than the average energy of the other water molecules, and then they turn to gas, carrying thermal energy away from your body and into the room.

I think i have solved it !!!!! Before anybody thinks he could send it in to get the 1000 pound before me,they do not accept any more entries ;)

First of all this works not in every case, it depends on the container (you will later see why). The container should be 4-10 cm diameter and at least 5 times as high, max 10 times as high. One must be nearly boiling, the other lukewarm.
Well if we have say 25° Celsius in one batch and say 98° Celsius in the other batch, round container 6cm diameter 42 cm high (filler till 40 cm) The hotter water evaporates faster and this evaporation leads to a loss of energy (quite a bit energy). Going from a liquid to a gas simply needs much energy. So this water cools quite quick. Now when it cools it gets heavier and sinks down. Now the next layer of hot water is on top, and the whole thing repeats with the next layer. It is pretty much the same principle as a lava lamp. There you also see that the difference in temperature doesnt need to be big, 1-2° is enough ^^.
Now the reason it outpaces the other liquid once the temperatures get closer and closer is that the water moves much faster in one container. You can also observe that in a lava lamp, it will go on for quite some time after you quit putting more energy in the container. So the one that started hot simply has constantly the hottest layer on top and exchanges the layers much faster. That way it has more evaporation at equal temperature then the other one. So while both may have the same average temperature, the hottest layer is hotter and the coolest is cooler in the container that started with hot water. So you still have more evaporation here

sorry for my bad english, i hope it is still clear what i mean

Last edited:
Well feyn, that certainly answers my comment in post 16,

If the two test vessels have a different temperature distribution they can have the same average temperature and still coll at different rate. It may be debatable whether the liquid can be represented by one temperture if it is not uniform but that would seem a good way forward.

Well feyn, that certainly answers my comment in post 16,

If the two test vessels have a different temperature distribution they can have the same average temperature and still coll at different rate. It may be debatable whether the liquid can be represented by one temperture if it is not uniform but that would seem a good way forward.
Hi studiot

well after thinking everything through that is the only explanation i can think of that makes any sense whatsoever. If you observe lava lamps,and meassure the temp difference there,you realize you get quite a current going with just little temperature differences ( 3° are already enough ). Once you have the current going it keeps going and going for quite long. But you need a relatively high start temperature to get it going in the first place, which is why i think either only the hot one has that effect at all, and the lukewarm has no inner current, or the inner current in that one is only very slow, that would have to be tested. That way the one with the current should be able to loose temperature faster when both are of equal average temperature.

K^2
Science Advisor
Lava lamp's heater has quite a bit of heat still in it when you disconnect the power. That's what allows a lava lamp to keep going for a while. The inertia of water in a container is absolutely minimal. Nowhere near enough for one fluid to pass the other in terms of cooling time.

You can verify this by keeping track of temperature in both containers, and when the warmer container gets close to catching up, drop some food coloring in both. You'll see nearly identical rates of convection.

I'm starting to think that the issue might simply be volume loss due to evaporation. Has anyone actually checked how much fluid remains afterwards? Even a small defect in mass would make a huge difference in freezing time.

I am starting to suspect it is not a single explanation at all, but the sum of various effects that summed up lead to this puzzling phenomenon ;) I will say more about that when i have more time