Empty
Occupied
Regular Cluster
Percolating Cluster
Interactive visualization of percolation theory and emergence of spanning clusters
Percolation theory studies how connectivity emerges in random systems. Consider a grid where each site is occupied with probability p. Neighboring occupied sites form clusters. As p increases, clusters grow and merge. At a critical threshold p_c ≈ 0.593, a spanning cluster suddenly appears that connects the entire system—this is a continuous phase transition.
Clusters are small and disconnected. The largest cluster size scales as O(1). No global connectivity exists.
Power-law cluster size distribution. Fractal spanning cluster with dimension 91/48 ≈ 1.896. Universal behavior independent of lattice details.
Unique infinite cluster exists. Largest cluster size scales as O(N). System is globally connected.
Near p_c, the system exhibits universal behavior characterized by critical exponents. For 2D percolation:
These exponents are universal—the same for all 2D lattices and even for continuum percolation.
Epidemic models use percolation to predict disease outbreak thresholds. Below critical infection rate, diseases die out; above it, epidemics spread.
Conductivity of composite materials with random conductive fillers. Percolation threshold determines when material becomes electrically conductive.
Habitat fragmentation and species connectivity. Below threshold, populations are isolated; above it, migration becomes possible.
Resilience of communication networks to random failures. Critical fraction of nodes that must fail to disconnect the network.
Percolation theory was introduced by mathematicians Broadbent and Hammersley in 1957 while studying gas masks with porous carbon filters. They asked: When do pores connect to form a continuous path? This led to the development of percolation theory, which became a cornerstone of statistical physics and the study of critical phenomena. The 2D square lattice percolation threshold was proven to be approximately 0.593 for site percolation.