Why does frost on a car sunroof form in crossing linear patterns?

[DaveC426913]

What is the physics behind this kind of frost, seen on the sunroof on my car?

Why does it form so linearly, even though every line has feathery offhsoots? (Why don't they all form radially, more like snowflakes?)

Why are lines able to cross each other with apparently no interaction?

I could see if these lines were being formed from pre-existing scratches in the windshield - that would explain both the above questions, since it would mean the structure of the lines has much more to do with the surfacfe and much less to do with the formation of crystals. But wouldn't the scratches have a more regular pattern, such as in line with the car's motion?

 

[lightarrow]

What is the physics behind this kind of frost, seen on the sunroof on my car?

Why does it form so linearly, even though every line has feathery offhsoots? (Why don't they all form radially, more like snowflakes?)

Why are lines able to cross each other with apparently no interaction?

I could see if these lines were being formed from pre-existing scratches in the windshield - that would explain both the above questions, since it would mean the structure of the lines has much more to do with the surfacfe and much less to do with the formation of crystals. But wouldn't the scratches have a more regular pattern, such as in line with the car's motion?

It would depend on scratches only, if the windscreen were perfectly clean; the process of crystals nucleation is complex, it could also be due to the presence of microscopic dust particles or glass imperfections but also on pre-existing (before using the car) microscopic scratches.

 

[Andy Resnick]

Dedritic crystals (like snowflakes) form when the energy balance between new surface area and phase change is negative, IIRC. That is, when it's energetically favorable to create new surface area (because the interfacial energy is low), a nucleating crystal will form a dendritic hanbit. There was a good experiment called "Isothermal Dendritic Growth Experiment (IDGE)" done on the space shuttle. A good discussion of the physics is here:

http://www.rpi.edu/locker/56/000756/

Now, on a scratched surface, the surface energy will be lower in some places than others (due to the scratch), and so a crystal will preferentially follow the groove.