new crystal-growth model

[phantom power]

(This might also have a positive impact on the manufacture of solar cells, if they can apply this insight to reduce the number of crystals in polycrystalline silicon.)

To the wonderment---and the befuddlement---of scientists, the patterns that form as plastics, metals and many other materials crystallize can vary incredibly, ranging from sea-urchin-like spheres to elaborate tree-like branches.

Now, Hungarian and National Institute of Standards and Technology scientists report in the September issue of Nature Materials* that they have developed a way to predict the polycrystalline microstructures that will form as complex liquid mixtures cool and solidify. Ultimately, the team's new simulation tool could help manufacturers of everything from plastic bags to airplane wings to design new products with improved strength, durability and other properties.

Images generated with the team's mathematical model match up almost feature for feature with the seemingly random crystal patterns formed in experiments as temperatures or other processing variables are modified. The model accurately predicts how both impurities (or additives) and process differences affect the sizes, shapes and orientations of crystals that form during the so-called supercooling process.

Findings based on the model indicate that instabilities along the boundary between liquid and solid areas during solidification effectively clash with the otherwise orderly process of crystallization. Tiny crystals-in-the-making move and position themselves along the growth front, assuming an orientation peculiar to the energy conditions at their location. Varying local conditions produce crystals in seemingly disordered arrays, accounting for the rich diversity of microstructural patterns.
http://www.sciencedaily.com/releases/2004/08/040831090728.htm

[bemildred]

1. Wonder how it fits in with making glass?

The story story a bit ago about making metal glasses
comes to mind, with enough disorder you get a glass? You can
get disorder by fast quenching or by adding "dirt", or both.

[phantom power]

2. as I understand it, a glass is not crystalline at all

A true glass has no crystal structure. It's actually a very viscous liquid.

Or, is it just very very multi-crystalline? I've seen 100-year-old windows that have flowed. They are thicker at the bottom than at the top.

[bemildred]

3. Something like that, a very slow liquid.

This piece, as I read it, says they have a model that based
on "impurities" and how fast things cool predicts the, ummm,
degree of crystallization, in the sense of how big the individual
crystals get. So it would seem at the extreme of fast quenching
and/or high impurities you get a glass. Now if you want to make
a glass, that could be useful.