Can condensation really heat your beer? If you've read our Quick Study, 'Condensation, atmospheric motion, and cold beer' in the April 2013 issue of Physics Today, you know the answer is 'yes.' Cold drink cans warm up significantly faster in hot, humid locations than in hot, dry locations—by approximately a factor of two in typical summertime weather conditions.
To reproduce the results of the experiments described in our Quick Study, or to see the effect of condensation on drink cans in your location, follow the experimental procedures described below.
But first, you might enjoy watching this video, which shows a slightly less scientific demonstration of how condensation can also help melt frozen drinks. It was produced in collaboration with the University of Washington department of atmospheric sciences outreach group.
The experimental procedure
We placed bottle tops, plastic snap-on tops that allow resealing of canned drinks, onto 12-oz aluminum cans filled with water. We then replaced the twist-off cap for the bottle top with a single-hole rubber stopper. As shown in the first photograph below, a digital thermometer was inserted through the hole in the stopper and an enlarged opening in the top of the can, so that the thermometer sensor penetrated well into the can's interior.
![]() |
The cans sat in an ice–water bath until they cooled to about 1?C. After selecting a can for an individual trial, we removed it from the bath, dried it off, shook it to mix its contents, and recorded its temperature. We then placed the can on a saucer capable of catching condensate that might subsequently run off the can's sides and weighed the can and saucer together using a precision scale.
For each experiment, we put the can–saucer unit inside an enclosed, temperature-controlled chamber for five minutes. After that time, we removed and reweighed the unit; the following photo shows the operation being carried out by undergraduate research assistant Stella Choi. The difference from the original measurement was due to the mass of condensate on the can and in the saucer. Finally, the temperature of the can was measured after its contents were shaken up.
![]() |
Results
The experimental procedure was repeated for a wide range of temperature and humidity. The following plot shows data from nine experiments, all taken at 25 °C and together covering a range of relative humidity. (The data shown are reproduced in the print Quick Study, along with those for the experiment run at 35 °C.) The filled-circle data points indicate the actual temperature change (δT) measured after five minutes. The open circles are an estimate of the temperature change from latent-heat release, calculated from the mass of condensate collected in the can and saucer. The condensate mass is multiplied by the latent heat of vaporization and divided by the heat capacity of the can to convert into a temperature change.
![]() |
Data from 18 experiments at 30 °C and varying relative humidity are plotted below.
![]() |
And here are data from 23 experiments at 35 °C and varying relative humidity.
![]() |
Finally, we present data from 24 experiments at 40 °C and varying relative humidity.
![]() |
The left edges of all the lines in the above plots correspond to no condensation and thus no latent heating. Condensation fails to occur when the surface temperature of a can warms above the dew point. Indeed, at the cutoff, the final temperature of the cans was within a degree or so of the air's dew point, calculated from the relative humidity and the air temperature. However, the temperature of the cans was about 1 °C less than the dew point. One possible explanation is that the cans' surfaces were slightly warmer than their contents. Or perhaps a small amount of water condensed on the cans but evaporated before we made our measurements.
The experiments were performed by undergraduate research assistants Stella Choi and Steven Brey. Galen Richards and Jaycyl Golding, high school students serving as Pacific Science Center Discovery Corps interns, worked on earlier versions of the experiments. Instrument makers Allen Hart and Steven Domonkos built experimental apparatuses. Funding was provided by NSF grants AGS-0846641 and AGS-1138977.