Hello, On Wed, Jul 1, 2009 at 3:59 PM, Daniel Wheeler<[email protected]> wrote: > > On Wed, Jul 1, 2009 at 7:12 AM, Ionut Vancea<[email protected]> wrote: >> >> Hello, >> >> just a short email to ask if you could find the nucleation process >> using CH example provided with fipy? > > Bit confused by the question. Do you mean "define" the nucleation > process? I'm not actually certain whether CH is a true thermodynamic > nucleation process in the sense a crystal growing would be requiring a > seed above a certain radius for growth. It may be that CH induces > separation for any finite sized noise, but I'm not sure (or if that is > even important for the definition of nucleation, maybe someone can > clarify). Regardless, I think that the initial noise distribution is > important for the type of CH process that you want to model. Anyhow, > in the example in the manual it is just a uniform distribution last I > checked. FiPy has other distributions and scipy probably does as well > if you want to seed it in a different way. > > Cheers > > -- > Daniel Wheeler > >
As I know, phase separation process can be classified as a nucleation process or a spinodal decomposition process. For example, quenches from the stable region (above critical point) to any point below the binodal curve will proceed via nucleation process, which is characterized by the random appearance of holes which will grow in time. The other possibility is quenches from a point above critical point to a point bellow the spinodal line, and that process will proceed via spinodal decomposition. A simple example of that can be seen here: http://www.fenics.org/wiki/Phase_separation_modelled_by_the_Cahn-Hilliard Cheers, Ioan
