A Links-Nodes-Blobs (L-N-B) model, based on the fractal and percolation con
cepts, is used to study the electrically conductive mechanism of conductive
filler loaded polymer composites. The change in the conductivity of polyme
r composites during the mixing process can be explained as the competition
between the breakdown of filler aggregates and the diffusion of ingredients
of matrix material and impurities onto the surface of the filler. The valu
e of the fractal dimension mu, which is the exponent in the power-law relat
ionship of the electrical conductivity sigma = sigma (o). (phi -phi (c))(mu
), is calculated as 1.88. This value is close to the values obtained direct
ly from experiments or from other simulations. The positive temperature coe
fficient (PTC) behavior in the conductivity of composite material is also e
xplained by this model as the breakdown of the conductive filler network. I
f the thermo-expansion induced strain is greater than the apparent onset st
rain epsilon (onset) = mQ+2 (G) over bar /2d (G) over bar. epsilon (b) of t
he L-N-B model, a strong PTC effect would happen.