snotitle.jpg (19016 bytes)

grow.gif (10037 bytes) Home
Natural Snowflakes
  --Photo Gallery I
  --Photo Gallery II
  --Photo Gallery III
  --Guide to Snowflakes
  --Snowflake Books
  --Historic Snowflakes
  --Ice Crystal Halos
  --Snowflake Store
Designer Snowflakes
  --I: First Attempts
  --II: Better Snowflakes
  --III: Precision Snow
  --Snowflake Movies
  --Free-falling Snow
  --Designer's Page
Frost Crystals
  --Guide to Frost
  --Frost Photos
Snowflake Physics
  --Snowflake Primer
  --Snow Crystal FAQs
  --No Two Alike?
  --Crystal Faceting
  --Snowflake Branching
  --Electric Growth
  --Ice Properties
  --Myths and Nonsense
Snow Activities
  --Snowflake Watching
  --Photographing Snow
  --Make Your Own
  --Snowflake Fossils
  --Ice Spikes
  --Activities for Kids
Snowflake Touring
  --Snowflake Hot Spots
  --Northern Ontario
  --Hokkaido, Japan (2) (3)
  --Michigan U. P.
  --California Mountains
Copyright Issues
Free-falling Snow
   ... Making snowflakes the natural way...

   Another technique for producing large numbers of snowflakes is to grow them as they fall freely through the air in a growth chamber.  Snowflakes made this way don't grow as large as the natural variety, however, since our growth chamber is not as large as a cloud.  But they are just the right size for many purposes, and we can photograph them when they fall onto a window at the bottom of the chamber.

Growth in a Convection Chamber
freefallx.jpg (8129 bytes)   We use a convection chamber to grow these crystals, shown schematically at right.  Basically it is just a cold chamber about a meter tall, with two containers of heated water on the bottom.  Convection mixes the water vapor into the cold air, creating supersaturated air for growing snowflakes.  We nucleate crystals by dropping a speck of dry ice in the chamber, or by rapidly expanding some cold compressed air inside the chamber.
   The crystals float until they grow to about 10-100 microns in size, at which point they fall to the chamber bottom.
Crystals Grown at Different Temperatures
Growth at -15 C.   The most dramatic snow crystal growth occurs at -15 degrees Celsius and at high supersaturations (>5 percent), which results in the growth of plates and stellar dendrites (see the Snowflake Primer). 
001ax.jpg (5103 bytes)38ax.jpg (2332 bytes)020ax.jpg (2743 bytes)34ax.jpg (1679 bytes)14bx.jpg (900 bytes)19ax.jpg (994 bytes)18ax.jpg (2519 bytes)15cx.jpg (2017 bytes)
20bx.jpg (4007 bytes)07ax.jpg (4379 bytes)39ax.jpg (2251 bytes)29bx.jpg (1333 bytes)018a1x.jpg (1533 bytes)06ax.jpg (1935 bytes)23ax.jpg (1577 bytes)19bx.jpg (1234 bytes)
t15mix2x.jpg (15529 bytes)The images above are only a small sample of the crystals we grew using the free-fall technique; the majority were simply measured and discarded.  The first image shows a crystal with a tip-to-tip diameter of 140 microns, and all the images on this page are shown at the same scale.  The image at right shows a mosaic of snow crystals grown at -15 C.
57ax.jpg (3104 bytes)17bx.jpg (896 bytes)    These crystals were all grown under quite similar conditions, although some grew for a longer period of time before falling onto the observation window.  The longest growth time was about two minutes.  Even the small variations in temperature and supersaturation within the growth chamber resulted in the great variety of forms seen.  This demonstrates that the final snow crystal shape is very sensitive to growth parameters, especially at -15 C when the supersaturation is high.
   We also grew many snow crystals at lower supersaturation levels (again at -15 C), and under such conditions the crystals nearly always grow as simple hexagonal plates, similar to the one at left.  Next to it is a photograph of a human hair at the same magnification.
Growth at -5 C.   Another interesting temperature for snow crystal growth is at -5 degrees Celsius, which for moderately high supersaturations (about 5 percent) yields hollow column growth (see the Snowflake Primer). 

58ax.jpg (3681 bytes)59bx.jpg (3109 bytes)48ax.jpg (1856 bytes)41ax.jpg (2009 bytes)46ax.jpg (1083 bytes)50ax.jpg (1323 bytes)59ax.jpg (1221 bytes)


Some of the smaller crystals grown at -5 C are in the form of nearly solid (hexagonal) columns, while larger crystals (or those grown at higher supersaturations) exhibit the distinctive hollowing of the basal faces.  This hollowing is a result of an instability in faceted growth, which causes the protruding edges to grow faster than the centers of faces; the same instability produces dendritic arms on plate-like crystals (see Snowflake Branching).  Several of the above crystals also show the twin-prism morphology, indicating that the initial seed crystal formed with cubic, not hexagonal symmetry (see Unusual Forms).
Growth at -2 C.   At -2 degrees Celsius snow crystal growth becomes plate-like again, although the growth rates are not as high as at -15 C (see the Snowflake Primer).  Thus the plates are smaller, and they show some qualitative differences.   Note, for example, that the larger crystals (at the far right below), which grew at the highest supersaturation levels, show distinctive rounded extensions.  These reflect the roughening transition that occurs near the melting point -- the ice surface becomes microscopically rough, so the crystal boundaries are no longer faceted.
26ax.jpg (1601 bytes)230ax.jpg (1556 bytes)35ax.jpg (1828 bytes)33ax.jpg (1568 bytes)32ax.jpg (1196 bytes)30ax.jpg (1220 bytes)30bx.jpg (1379 bytes)31ax.jpg (1178 bytes)


The Hardware
lab1x.jpg (4877 bytes)lab2x.jpg (2534 bytes) At the far right we have a picture of the growth chamber, which is covered with styrofoam insulation and topped with a high-intensity lamp.  The refrigerator is at the lower left in the image.  The principal players this year -- Audrey Chng, Ken Libbrecht, and Haitao Yu -- are shown in the near image.
   Since the crystals grown in the free-fall chamber were quite small, we used a high-quality microscope to observe them (N.A. = 0.28).  Most of the crystals photographed above were not much larger than the diameter of a human hair.
   In addition to taking pictures, we used this apparatus to make a number of precise measurements of the growth rates of ice crystals under different conditions.


Return to was created by Kenneth G. Libbrecht, Caltech
Comments?  Send an e-mail....
page views since February 1, 1999