Scribd
Upload
12 views9 pages

Lesson Percolation

Percolation theory describes the properties of clusters in a random graph as the probability of occupying sites, p, varies. There are three key points: 1) There exists a critical probability pc, called the percolation threshold, below which only small, disconnected clusters form and above which an infinite, percolating cluster emerges that spans the entire graph. 2) At the critical point pc, the size of the largest cluster scales as a fractal dimension D. Above pc, the largest cluster size grows linearly with the system size, while below it grows logarithmically. 3) At pc, the percolating cluster is scale invariant, meaning subclusters have the same statistical properties as the

Uploaded by

Laulesh Padhan
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
12 views9 pages

Lesson Percolation

Percolation theory describes the properties of clusters in a random graph as the probability of occupying sites, p, varies. There are three key points: 1) There exists a critical probability pc, called the percolation threshold, below which only small, disconnected clusters form and above which an infinite, percolating cluster emerges that spans the entire graph. 2) At the critical point pc, the size of the largest cluster scales as a fractal dimension D. Above pc, the largest cluster size grows linearly with the system size, while below it grows logarithmically. 3) At pc, the percolating cluster is scale invariant, meaning subclusters have the same statistical properties as the

Uploaded by

Laulesh Padhan
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
You are on page 1/ 9

PERCOLATION

Mario Nicodemi

Complexity Science & Theor. Phys., University of Warwick”

Topic:
• Basic concepts in percolation theory;
 Percolation
p=30% p+5% non percolating
cluster

percolating cluster

p=35% p+5% p=40%


 Normal and percolanting clusters

f Fraction of space occupied by clusters


1 percolating
cluster

non percolating
clusters
0 pc 1 p
 Critical point

L The fraction of filled


sites is named p.

p<pc p~pc p>pc


The percolation probability, P (p), is defined as the probability that a “cluster”
of filled sites spans the whole lattice.
P(p)
If L → ∞, P (p) = 0 ∀p < pc .
1
The critical threshold pc is known as
L1 < L2
percolation point. L3 =

0 pc 1 p
 Percolation threshold

p=0.57 p=0.59 p=0.61


 Percolanting cluster
The “mass”, M , of percolantion cluster is defined as the total number of sites
belonging to it, or, equiv., its area. M scales as:

• if p > pc: M ∼ L2
• if p = pc: M ∼ LD (e.g. D ' 1.89 in d = 2)
• if p < pc: M = 0

ln(M)

ln(L)
 Summarising P(p)
• The percolation probability P (p) is 1
L1 < L2
- for p < pc : P (p) = 0
L3 =
β
- for p ≥ pc : P (p) = (p − pc )
E.g., in d = 2 we have β ' 0.14, and
on a square lattice pc ' 0.593. 0 pc 1 p
• The number of sites, M (L) (Mass), of the largest cluster (which for
p ≥ pc is the percolating one), in the limit L → ∞ is:
- per p < pc : M (L) ∼ ln L
- per p = pc : M (L) ∼ LD ln(M)
- per p > pc : M (L) ∼ Ld
Note that for p = pc , the percolating
cluster is a fractal. E.g., in d = 2
it is D ' 1.89. ln(L)
 Cluster size distribution
The probability, to find a cluster with size s (i.e., with s sites) is:
• for p < pc:
n(s) ∼ s−τ exp(−s/s0) log(n(s)) p<pc
0
• for p = pc: 10 p=pc
n(s) ∼ s−τ p>pc
with τ = (d + D)/D
• for p > pc: −1
−τ
n(s) ∼ s exp(−s/s0)+ 10
+δ(s − sM )
s0
−2
10
log(s0) log(sM)
log(s)

0 pc 1 p
 Scale properties
• Percolating cluster is characterised at pc by a very important feature,
named scale invariance:
if you take a subset, S 0 , of the percolating cluster S , and blow up with
a scale factor r (i.e., multiply its points coordinates by the factor r)
you get a set rS which is statistically congruent (“equal”) to the
original set S .
For such a reason, the percolating
cluster is named self similar.

You might also like