Study of Superexponential Growth of the Mean Particle Flux by Monte Carlo Method.

A comparative analysis of two algorithms ("by particles" and "by collisions") for estimation of the weighted mean particle flux is made on the basis of a test problem solving for a single-speed particle propagation process with isotropic scattering and multiplication in a random medium. It is shown that the first algorithm is preferable for a simple estimation of the mean flux and the second one fits better for assessment of the parameters of a possible superexponential growth of the flux. Two models of the random medium are considered: a chaotic "Voronoi mosaic" and a "spherically layered mosaic." For a fixed mean correlation radius, the superexponential growth turned out to be stronger for the layered mosaic. [ABSTRACT FROM AUTHOR]