Entropy in the Brazil Nut Sorting Effect

[Tahmeed]

In Brazil Nut effect /Granular convection the large grains move upward and the smaller ones go downward. This sorting is supposed to reduce the multiplicity of this system. But according to the second law of thermodynamics, entropy and multiplicity of the system should increase.

I am looking for a detailed and intuitive explanation of this.

 

[BvU]

Consider what you define as 'the system'. Does it have a fixed energy ?

 

[A.T.]

But according to the second law of thermodynamics, entropy and multiplicity of the system should increase.

What does it say exactly?
https://en.wikipedia.org/wiki/Second_law_of_thermodynamics

 

[Chestermiller]

Entropy is a molecular scale phenomenon. It is not expected to apply to macroscopic systems, such as how you organize your library (or how Brazil nuts gravitationally segregate).

 

[BvU]

I am looking for a detailed and intuitive explanation of this.

Google is your friend. There's a lot behind pay walls, but I liked browsing this one, so: thanks for bringing it up !

Consider what you define as 'the system'. Does it have a fixed energy ?

My reason for putting it this way: I figured potential energy is won when interstitial void is reduced (closer packing). This energy is dissipated and can't be reconverted. But Chet's comment is quite valid although my own library definitely obeys the second law.

granular materials are a system out of equilibrium at almost any level