Effective conductivity of many-component composites by a random walk method

An extension to the many-component case is given for a random walk algorithm which was recently introduced to calculate the effective conductivity of random mixtures with two finite conductivity components. The many-component algorithm has been tested for the case in which the conductivity distribution is generated by taking N-component conductivities and assigning them at random, with equal probability, to the sites of a lattice of size 30 3 . Simulations have been done with N=10, 100, 1000 and 27 000. Numerical results obtained by using the proposed random walk algorithm are shown to be in close agreement with those obtained using the conventional finite difference method.