By the way, regarding the "full bottle of air ascending from depth" experiment, here's a little background from a home soda maker (I wear many hats).
There is a note in the article "
Carbonating at Home with Improvised Equipment and Soda Fountains" which says, "Compare this to Coca Cola's standard of 3.7 volumes for their bottled product." Let's take that as our benchmark for fresh carbonated beverages. The unit "volumes (of CO
2)" is not common outside the beverage industry. A carbonation level of one volume of CO
2 basically means that each liter of soda has one liter of CO
2 dissolved in it. (It's just a unit of convenience. :biggrin
So, how do you convert volumes of CO
2 into something more useful? Well, you could use some lovely complicated formulas, but most of us just use a simple solubility table... and I just happen to have a
CO2 solubility chart (PDF) right here. You probably don't need to be told that CO
2 solubility in water is a function of temperature and pressure. Solubility (in volumes) is the body of the table, temperature (in degrees Fahrenheit) is the vertical scale, and pressure (in PSI, gauge) is the horizontal scale.
So, if we know that our bottle of Coke has 3.7 volumes of CO
2, we can see what pressure will result from any given temperature. For example, at an almost ice-cold 34 degrees Fahrenheit, the pressure will be something like 19 psi. If we happen to be ice diving with our Coke, and we open it at 34 degrees Fahrenheit and 45 feet depth, there will be no fizz at all.
Anyway, so what if we leave the Coke in the car while we go get tanks filled on the way home? Let's say that the Coke in the bottle gets up to a not-very-delicious 90 degrees Fahrenheit. According to the table, then, the pressure inside the bottle would be around 73 psi (gauge). In my copious experience with hot soda (hey, this is Louisiana), I've never had a bottle burst due to being left in the car. At a mere 90 degrees Fahrenheit, the pressure is equivalent to filling the bottle with air and sealing it at about 164 feet (roughly 50 meters) and sending it to the surface.
A quick Google search ([google]burst pressure soda bottle[/google]) mentions (among many other things) a 120 psi burst pressure for a two liter soda bottle. That would be somewhere around 270 feet or about 82 meters depth for the fill-release-and-boom thing to work. (I haven't put a pressure gauge to a bottle while doing the "dry ice bomb" experiment, as that is *quite* a dangerous an experiment to begin with, but it'd be interesting -- although the thermal effects may be a confounding factor.)
Oh, and in case any of you want to use soda as a fun diversion, here's a little table (made from the 3.7 volumes assumption). For a given temperature, if you open the soda deeper than the listed depth, it should not fizz (regardless of how much you've shaken it):
Temp(F)|PSI(g)|Depth(fsw)
85|67|150
80|62|139
75|56|126
70|50|112
65|45|101
60|40|90
55|36|81
50|32|72
45|27|61
40|23|52
35|20|45
(Oh, and as for my sodas... I carbonate them to about 5 volumes, which is sting-your-mouth fizzy. Anything else tastes flat to me now.

)