What did we learn?
The Chemex Cozy works.
As you can see on the chart below, the cozy kept the liquid warmer throughout the test period, even though (somewhat mysteriously) the initial temperature at 20 seconds was a few degrees cooler when the cozy was on:
At 30 minutes with the cozy, the liquid was still about as hot as it was with no cozy after 10 minutes.
It's also worth noting that the one big drop in temperature came at the moment I had trouble with the thermometer. Further study -- and a new battery for the Thermapen -- may be required. It may be that the more gradual decline seen in the first 10 minutes with the cozy (drops of 9 and 4 degrees, compared with 24 and 13 degrees in the control condition) would hold for longer. Other explanations for the discontinuity in the rate of cooling are welcome.
Praise be to the Chemex Cozy! And thank you, Grandma Joy! The cozy looks great, and it works like a charm. Mornings will be better -- or at least warmer -- from now on.
The Chemex Cozy works.
As you can see on the chart below, the cozy kept the liquid warmer throughout the test period, even though (somewhat mysteriously) the initial temperature at 20 seconds was a few degrees cooler when the cozy was on:
At 30 minutes with the cozy, the liquid was still about as hot as it was with no cozy after 10 minutes.
It's also worth noting that the one big drop in temperature came at the moment I had trouble with the thermometer. Further study -- and a new battery for the Thermapen -- may be required. It may be that the more gradual decline seen in the first 10 minutes with the cozy (drops of 9 and 4 degrees, compared with 24 and 13 degrees in the control condition) would hold for longer. Other explanations for the discontinuity in the rate of cooling are welcome.
Praise be to the Chemex Cozy! And thank you, Grandma Joy! The cozy looks great, and it works like a charm. Mornings will be better -- or at least warmer -- from now on.
I commend your thoroughness. You are now ready to homeschool your children. That said, I will need to send an independent lab tech and coffee taster to verify your results. Could we set up an appointment the day before Thanksgiving?
ReplyDeleteI'll have my people get in touch with your people.
P.S. Of course it's hot!
ReplyDeleteAhem. The rate of cooling is not constant -- according to my observations and calculations (and some dude named Newton), the rate of change of temperature is proportional to the difference between the temperature of the object and that of the surrounding environment. Let's just call this "Newton's Law of Cooling", just for the hell of it.
ReplyDeleteWhat you'd expect to see in your graphs is a curve that is steepest at t=0 mins and flattest at t=30 mins, with an exponential decay in between. The temperature of the coffee will approach the temperature of the environment asymptotically -- steep drop in temperature at first (when the differential between the liquid and the environment is greatest, then leveling out as it cools toward the ambient temperature. The curves for the GJoy enabled and non-GJoy enabled Chemex will have the same basic shape, but the curve for the uninsulated carafe will be steeper.
One reason that you aren't seeing this in your empirical data is that you haven't graphed it against a linear X axis... meaning, the intervals between the data points aren't spaced out correctly - the horizontal distance between the datapoints at 5 minutes and 10 minutes should not be the same as the horizontal distance between the datapoints at 10 minutes and 20 minutes. Correct this, and you'll be closer to Newton.
All this to say, nice work Ma!
True, the heat transfer should be greatest when there is the biggest difference between the carafe and its environment, and this is what we see in the control condition: fastest drop in the first 5 minutes, slowing in the next 5 minutes, and so on. This also makes me think the thermometer compromised the data, since in the Cozy Condition, we see the steepest decline after 10 minutes. As Newton himself no doubt would have asked, WTF? (Or maybe, W dT/dt F) Also true that the X axis is distorted -- an excellent observation that reveals the limits of my excel skills and a sin that probably requires me to return my Tufte books -- so the slope of the curve does not reflect the actual rate of cooling. However, since the two curves are distorted in the same way, they should remain comparable at least for the purpose of watching the relative decline between the two states.
ReplyDeleteOn the other hand, the absolute data is clear: GJoy keeps it hot.