come on let's go and play....
Well .. maybe a snowflake or two rather than a snowman.
Years ago C.L. Strong's Amateur Scientist column in Scientific American talked about homemade ice crystals. Homemade apparatus for carefully controlling temperature and humidity so you could study their growth. These pieces were great in that they had appeal to a wide audience. Remarkably just enough to suggest how you might build something sophisticated along with the physics or chemistry of what was going on. Enough to pique your curiosity to the point where you wanted to do something. You had to "lean forward." It has been said this column excited more young people about physics and chemistry than Sputnik and the space age.
The serious apparatus described was too expensive and challenging a fourteen year old, but it mentioned the right points - namely that you want to set up a temperature gradient between a warm damp region and a cold one. Place a pin or a string .. anything very thin, between them and, at the right temperature and humidity, ice crystals will form. Strong suggested a damp cloth positioned above a chunk of dry ice to get going.
That was enough for me. My father had a serious Albertson's ice cream addiction which required the regular purchase of chocolate chip or rocky road ice cream in the three gallon cylindrical containers. I rescued an empty, cut a slit down the side and created a tube about five inches in diameter .. the chunks of dry ice from the Talcott's were six inches on a side .. I had a vertical viewing port about two inches wide that was covered with a thin flexible plastic sheet (why I didn't make the cylinder out of that entirely is beyond me ... my style probably dictated an ice cream empty) A thin piece of fish line (any respectable house in Montana had a wide variety) ran vertically from a damp sponge towards the dry ice. A sinker at the bottom keep it pointing down. It took some fiddling around .. the airflow had to be very smooth .. but within an hour I had distinct types of crystals forming at different temperature locations along the line.1
I had been thinking about suggesting this as a neat family experiment and played around with something not very elegant a few years ago. I needed a better approach but had more or less given up on it when I caught a Science Friday interview of Ken Libbrecht. Ken probably knows more about ice crystals and snowflakes than anyone else on the planet and is something of a legend. When he suggested using a 2 liter softdrink bottle I had one of those slap the hand on the forehead moments. Of course - now it's simple.
If you can find dry ice, and that's not terribly difficult (I asked Siri in Manhattan and found a half dozen locations), you're almost set. It is beautiful to watch and you can go much deeper. Those of you so inclined can set up a camera (dark background with side illumination and some magnification) and make beautiful time-lapse movies. You can vary the conditions and see different types of crystals form. You'll develop a sense of how shape depends on temperature. And the basic physics is unsolved so far for those who want to really drill down. (hint.. you'll need to build a someone more sophisticated apparatus to ask that class of question properly). You can even see, rather dramatically, that warm flows to cold rather than the other way around (cold is really heat leaving a warmer object... so you don't get cold... you lose heat)
They even made a video which is much clearer than any written description I could provide.
Really fun for the family - way better than TV, although it's a bit inefficient for making snow to build snowmen or women.
1 One of the questions I had was how to accurately measure the temperature at any point. That turned out to not be easy - at least on my budget - as I didn't know about thermocouples. I tried to figure it out with math, but didn't have enough. That became one of the drivers that brought me to partial differential equations. But in the end ground truthing with accurate measurements is better.