Frozen Soda Can Did Not Burst

My WAG is that the crushing of this can is what damaged the temperature sensor.

Ooh, concentrated essence of The Pepsi Challenge.

Yes, I'm sure the can has a perforation and it may have a leak that is visible undetectable.

The limited available data suggests to me that there was no leakage from the can.

.

The OP says that the can remained sealed, but beyond that the OP notes that all the cans remained sealed.

.

That all the cans remained sealed is an important distinction, because it indicates there was likely no evidence of leakage in the machine (not just that the OPs inspection of the can alone made him think that it remained sealed).

.

There is a good chance that a portion of any leaked material would have frozen and been very noticeable when removing the cans and inspecting for leaks.

.

I am also dissuaded of the pinhole argument by its complexity. The argument's need for the pinhole to either only transmit fluids but not gasses, to reseal itself, or to act as a one way valve strikes me as unnecessarily complicated.

Yes, I'm guessing at it too......

So has the can been weighed and compared to others, or leak tested as others have suggested?

Here's some experimental data on the internal pressure inside a soda can, unfortunately at room temperature. Perhaps someone can take the appropriate gas laws and calculate the pressure given the conditions under study here. Using that as a starting point I assume the pressure exerted by the expansion of the freezing soda can be calculated along with the force necessary to change the bottom from concave to convex. That will give us a change in volume and a handle on the forces involved with cycling the temperatures. I'd do this but this is way out of my area of expertise!

Who wants to step up and solve this mystery??

WWhat happens if you put the can in near boiling water? Does it bubble?

Can is empty there is nothing left probably by the time it melts.

For me it looks like this can was squeezed by the other expanding cans and most likly was one of the later one added so containing still liquid coke which was squeezed out like a lemon.

Is there a spillage in the vending machine?

I don't think developing a pinhole leak is sufficient to explain the results.

.

The can was not dipped into a freezing substance that would be needed to freeze plug the pinhole leak prior air reentering the can as the contents cooled.

.

The can was cooled down slowly by the refrigeration system along with all the other cans.

.

Partial filling of the can as an explanation is consistent with the damage evident, including the bulging bottom. Pressure from the liquid inside could still be enough to distort the bottom of the can as it expanded through freezing, especially if the can were laying on its side as it is shown in the picture.

Notice how the bulge is more pronounced closer to the lower surface.

Also notice how the bottom of the can is not square with the sides. It appears to be drawn in, exactly as it would if buckling from higher pressure from the outside, but supported by expanded ice on the lower section...which would require less than a typical fill in the can.

.

.

If the under-filled can explanation is valid, side of the can that is not visible in the picture should not be as drawn in as the side we can see. Also a lack of any indication of impact would also support this explanation.

Beats me.

It seems like one under filled can on a production line would be highly unlikely.

My pinhole theory came from a story a buddy of mine told me, about a six pack of beer that he left in his basement for years, and despite no obvious holes, most of the liquid was gone from the cans...no signs of leakage either.

So several of the cans all developed non-obvious pin hole leaks sufficient to drain the contents ? That is odd.

It seems like draining beyond the initial pressure release would require at least two pin holes....

Perhaps your friend had a teenage son that might have had access to the basement during that time?

I agree with you. I know soda's (being from the South, we would normally call them the generic 'coke;, but since this is obviously Pepsi, why bring an uninvolved party in?) over a period of time will go 'flat', which is accelerated by high temperature. I think this is caused by slight permeation of the CO2 through the seam. If this can experienced high temperature, that could have caused the pressure to get high enough to bulge the bottom and possibly bleed off the carbonation. Then, when the temperature dropped, the can collapsed. Granted, I've never seen a bottom bulge from high temperature, but I've never exposed a cokePepsi to more than about 100degF. But it seems possible. Maybe OP would like to conduct an experiment.

Even with the bottom bulged out, it doesn't look like the amount of liquid normally in a can would fit in the crushed can.

An accurate weight scale could easily settle the debate. S.M.

That is a good idea.

It sprung a leak. The liquid evaporated and condensed on the coils. There is probably a sticky residue on the floor of the machine.

Looks like the bottom side which normally is inward spheric, went outward during a heat related, high pressure phase on an overfilled or dented (or other flaw baring) can. On a following (cooling?) low pressure phase, the can imploded. All you got to do is investigate the reason for that heating cycle. Hint: Global warp... er... warming, is out. S.M.

I believe it is physically impossible for a full can of soda to shrink when frozen, having had the experience of freezing soda cans. Are you quite certain that it didn't have a pinhole in it and that some of the soda didn't leak out before it was frozen? Another theory I have is that the formula was not complete in that the carbonation was not added prior to the can being sealed. Hey, maybe the machine that was supposed too fill it up, screwed up! I've also encountered cans that were sealed, but not full of fluid.

..I do not know the freezing point for Pepsi but I suppose that at at 16 F it has frozen. Based on this assumption, the volume of the liquid inside has increased causing the expansion of the can bottom as can be seen in the picture. Thus, physical volume of can has increased.

When temperature has gone back to normal liquid volume was decreasing thus causing inside pressure to drop. Since we may assimilate the can with a thin wall tube with reinforced end rings , then we suppose that the thin side walls - which are not at all stiffened- have collapsed/buckled inwards due to the de-pressure stress.

...then why did this not happen in any of the other cans?

I suggest that some of the liquid escaped through a pinhole when some of it froze and expanded. Then when the frozen liquid thawed and contracted, the pinhole became sealed by some mechanism and so the can had tocollapse as shown in the pic. I would ask Pepsi and, if they cannot give an answer, get them to talk to their can manufacturers. There will be a simple answer and the can makers will almost certainly have it.

As the temperature falls below 0 deg C, the carbon dioxide in the liquid solution of the can may contract more than the liquid contents expand with the results experienced.

That still fails to explain why only this one can exhibited the particular behavior.

Cant trust the Dioxide! Never! Its a shamble!

Dang Vampires!

Possibility:

If this can was found empty (i presume from the final shape), then:

- On cooling down , the CO2 dissolved in the liquid, will escape and increase the pressure in the can, (temperature and pressure not enough to liquefy the CO2). This will explain the bulging at the bottom.

- Then, a leak could have been formed (pin hole ?), where the liquid was pushed out so that a vacuum is formed inside. This will buckle the can as shown.

Sure. As the machine was cycling, this can hit the temperature sensor causing: 1) the sensor to fail and 2) piercing a small hole in the can. Some of the can contents leaked out resulting in an air space that caused the can to collapse as the internal fluid froze. Send the defective sensor to the lab to check for Pepsi residue on the tip.

"Why is this the only can like this?" is exactly the right question.

.

One point of correction, though: the OP does not say that none of the other cans were bulged, only that they remained sealed.

The OP does hint that the other cans were bulged when it is noted that this particular can 'went the other way'.

.

I find the resealing pinhole hypothesis highly improbable based solely on the level of complication.

.

The simple easy and more probable answer is that the can was under-filled. The can was laying on its side. The freezing liquid bulged the lower portion as displayed in the picture.

The differential between the lower gas pressure inside and the greater atmospheric pressure outside caused the can to buckle, slightly pulling the bulged can bottom toward the can top and pressing the thin walls inward in the places the frozen liquid was not occupying. That is all consistent with what can be seen in the picture.

That's a SQA (Statistical Quality Assurance) question. After getting the right amount and quality into the can, the sealing of the product is of utmost importance. Very rarely will an improperly sealed can get through the canning line, the likelihood of two or more in the same batch is very remote unless there's something wrong with the sealing equipment. So this one can was at the edge of the spec, the others were more in the middle, and so weren't affected.

Maybe more would have had the same failure mode if the temperature fell to zero and/or there were repeated freeze/thaw cycles. The fact that only one failed was a testimony to the SQA practiced by the canner.

Sorry but I don't buy that. The idea that this solitary failed can would be able to hold enough pressure to invert the domed bottom during heating. Then have the seam act like a pressure relief valve to lose product. Then to have a run away freezing process to drop the internal pressure with this being the only deformed can seems very implausible to me. During the heating stage I would expect a few of the other cans to have a bulging bottom that would wedge themselves in place. Since the OP did not inform us that other cans were stuck in place to corroborate this scenario, I believe the only thermal cycle must be the solitary run away freezing cycle.

My expanded WAG is that this specific can was original under filled or developed a puncture when the vending machine was filled. The CO2 pressure would drive product out of the damaged can to the level of the puncture. Then gas pressures inside and outside would come to an equilibrium.

Sensor failure now occurs making for a runaway cooling cycle. The under filled unpressurized can now has considerably less mass than the other cans so the remaining liquid in this can freezes first. The freezing water in the product expands and plugs the leak. Internal gas pressure now drops. The can crumples as icing inside the can continues. The freezing fluid in this horizontal can that's between the inward dome and the sidewall expands and puts tremendous pressure until the dome pops into an inverted (outward) dome. The remaining pressurized cans are just getting slushy when a service technician (OP?) arrives.

A very interesting hypothesis...now my fun to poke some holes

If the expanding ice plugged the hole, the temperature of the gas/vapour in the can was at about 0°C. The lowest temperature was 16°F (-8.89°C). I don't have the time right now to do the calculation, but I don't think dropping 9°C is enough to compress the gas/vapours enough to crush the can.

I would tend to agree with the thought that the can was not filled completely to begin with. I have seen cans of fizzy drink completely empty (never beer though!).

The hole in my argument is that buldged bottom. Would 120°F be enough to expand the air in an empty can to make enough pressure to cause that?

I also like what someone wrote about the hot can leaking, sticky residue sealing the hole then freezing. Why this can and not the others? A flaw in the metal, a scratch in the surface making a force concentration point?

Drew K

Reread our sparse information from the OP. The failed sensor reported 120°F. This lead to the run away cooling cycle. We do not know what temperature range this beverage was exposed to. I believe the two plausible scenarios start with a room temperature as this can gets delivered. Then depending on when the restocking delivery of this can happened in comparison to the temperature sensor being damaged this can then dropped to the run away temperature of 16°F in one or two stages.

As I said earlier, without any feedback from the OP we will only be able to make a batch of WAG here. A WAG can be fun to make though.

Could it possibly be that this can was not full to begin with, as I've stated previously, and that the other soda cans, in expanding, pressed against this one causing it to bulge only on the bottom and crushing the sides?

Thats what I think too!

Here's more info on the aluminum can, from Scientific American no-less, including a failure rate figure of 1 every 50,000 +/-. After reading about all the science that goes into their production it seems unlikely that the metal is still ductile enough to stretch and shrink, more likely deformation is permanent, so the "pinhole-reclosing" scenario probably takes a second seat to the "leaky-can" scenario.

Looks like we'll not have the honor of an OP reply. And since I am the only one that sees the top and bottom deformation as just about consistent with the side can compression, I quit! Pinhole reclosing or leaking can my b*t (LOL) S.M.

You see 'the top and bottom deformation as just about consistent with the side can compression'?

.

The main things that suggest to me that the buckling on the top (as it is oriented in the picture) is more extensive than on the side we can see (and the bottom which we cannot) are:

.

1. The bottom of the can (this time the part of the can designed as the bottom, the portion to the left in the picture) is not square with the top of the can. It is clear that it has been pulled significantly towards the top of the can on the portion viewed in the top of the picture.

The bottom of the can had to either be pushed out more at what it in the lower portion of the picture, or was drawn in the portion that is in the upper part of the picture, or some combination.

.

2. The shape of the can wall is very different in the lower portion of the picture than in the upper portion of the picture.

In the upper portion of the picture, hard creases appear and there are indications the wall was drawn in severely enough that portions of the wall interior pressed against other interior portions.

In the lower portion of the picture, the deformation has not created such hard sharp features, the indentation is rounded.

.

.

Are you seeing any of that differently?

I suspect most of us agree that if a sealed can with only a small amount of fluid in it were to freeze, the result might look like this. If you don't believe that one way leaks are a fact of life ask a clam, an oyster or think about how a harmonica works. In fact, the easy open top of these cans looks a lot like a harmonica seal. The liquid ran out and air could not get in to replace it so the differential pressure upon cooling crushed the can. I bet if you installed a one-way valve in a soda can, heated and then froze the can, you would replicate the result.

Of course you can replicate (not too closely) the result the way you say, but NOT only that way. Do you have an idea of the pressure needed to invert the curve of the can bottom dome? The can actually survived that pressure and wasn't beheaded ONLY because it was mostly slowly applied hydraulic pressure, because of the formed ice volume expansion, due to the runaway refrigeration and much less from the separation of Co2 at freezing. (Co2 and most gasses solubility in water based liquids peaks at near freezing point, and then goes near zero) and probably because the can was specifically designed to act that way. Now why one and not the others, I lean mostly towards possible overfilling, to have that kind of expansion, or a defect of the specific can, like a dent, that initiated the expansion at lower pressure. The implosion was at a second stage, after the ice started to melt. Near that point Co2 was diluted again in the liquid, so the pressure dropped, but this time we have a much larger can volume, and a possibly dented can that could not withstand much (NEGATIVE this time) pressure, so imploded randomly. Well not really randomly, but the final shape is determined by chaotic equations not really meaning much. It's just that you can't expect a very symmetric collapse of a thin wall container when pressure is from outside in. That also explains the points of TINAC's post 42 above, not to give answers twice. S.M.

I suspect that this was a stale (out of date) can that lost its carbonation, or was never filled properly at the filling line. With very slow loss of pressure, there will be a point reached where at normal vending temperature, the can is at external ambient pressure. Upon freezing temperatures being reached, the head space goes negative pressure which these can are not designed for. As the liquid finally freezes solid, the only place left to expand is to the domed (concave) portion of the can bottom, then the bottom becomes convex.

If never filled properly, then even if the carbonation was present, there would be excessive head space, and with increased solubility of carbon dioxide in the cold liquid, at some point the pressure went negative, same result upon final freezing.

Sorry to have not got back with more information. I have been in a mountain cabin with no electricity, phone, internet or other modern conveniences. In answer to some questions asked; The can was not out of date. No other cans were affected. The can is question weights 304 grams, the other diet pepsi's weigh 371 grams. There was no evidence of leakage on the bottom of the machine. It was very carefully checked and cleaned inside and out. When the freezing occurred, the machine was shut down and all the soda remained inside for about 3 weeks while waiting for a new part. When the new part was installed, I removed every can and inspected them one at a time to be sure there were no bad ones. The machine was almost empty at the time of the problem, so new cans were added following the incident. At least one of each flavor was opened and tasted to be sure there was no loss of flavor or carbonation. The damaged can was on the top of the stack. Any other questions I would be glad to answer. BTW I'm done with the recluse stuff, so I will be available for answers.

.

.

.

.

If that is the case, then you should look toward the graylings. They seem to be responsible for every unexplained mystery on television these days.

Was there anything special about the location of this can? Could it have been in the path of the blower comming from the chiller?

Can we get some more pictures?

Drew K

It seems pretty clear from the additional information provided that the can was simply under-filled. The can was missing probably a little less than 20% of the liquid it normally would have.

He could do a volume analysis to compare the difference in weight to the difference in volume. the easiest way would be to fill a container full and place the can in then measure the water that spills out, do the same with an undamaged can to get the volume difference.

I know the people that fill these machines do it quickly and sometimes with 2 cans in one hand but I would think that I would notice if a can were underfilled.

Drew K