Modeling a crack in a material.

[nukapprentice]

I was wondering if anyone has ever tried to model/fit a crack on a material surface with a Weierstrass function.

 

[nukapprentice]

I'm also wondering whether a crack exposed to damage becomes more jagged over time or more smooth? Is there a process that reverses the other one, i.e. from being more jagged to smooth, vice/versa?