Why does ice form in straight-line patterns on car windows?

[alexbib]

The last few days have been very cold and when I come back to my car, water has condensated (to solid phase) in the windows. I have noted that the ice seems to form in lines that cross each other (forming seemingly random patterns made of almost perfectly straight lines) Of course there is some condensation all over the windows, but the ice in these lines is much concentrated. If you know what I'm talking about, do you have any idea how these ice lines are formed?

Thanks,

Alex

 

[russ_watters]

I don't remember the chemistry term for it (seeding?), but the ice grows in crystals. It attaches easier to other ice than it does to the windshield. So when one crystal starts, the water vapor in the air will attach to it, making it grow.

 

[alexbib]

yeah, I thought about that, but why does it not form spots of ice rather than lines? Why is it easier to condensate on the axis of the line than in a random direction next to it?

 

[alexbib]

come on guys, I'd really like to know...

 

[jago25_98]

I don't really understand it. I actually found this thread search for an answer.

I'll share what little I know.

I think that actually there might be a lot a freezing pattern can tell us.

I know there's a patent on using freezing to remove nuclear waste from water. I can't remember what the pollutant name was though... it ended in `ium` and started with a D I think. Duerturanium isn't it... hmm...
basically some ice freezes quicker than others.

... anyway, I have been playing with freezing water and tasting the results. It certainly makes a difference. I noticed by looking in the bottle there is dirt on the surface of the ice.

I'm sure there is a word for the study of freezing water (what's the word for liquid to solid state change?). If we can find the patent I'm looking for there will be words in there that we can use to help us search for more info.

In the mean-time, here's this;
http://www.ncbi.nlm.nih.gov/pubmed/15247390
and this:
http://www.sciencemag.org/content/327/5966/672.abstract

edit:: sorry, totally bumped this old thread!