A computational algorithm for random particle breakage

Abstract

Random breakage can be defined as the breakage patterns independent from the stressing environment and the nature of the broken particle. However, the relevant literature studies give contrary evidence against random breakage of particles. A simple way to detect random breakage is to evaluate the fragment (progeny) size distributions. Such distributions are estimated analytically or through numerical models. The latter models generally treat random breakage as a geometric statistical problem with prior assumptions on particle/flaw geometry and external stressing environment, which may violate the randomness of the breakage process. This study presents a random-breakage algorithm that does not require such assumptions. The simulated progeny size distributions were compared with the experimental size distributions by impact loading (drop-weight) tests. Random breakage events should yield number-weighted size distributions that is fitted well to the lognormal distribution function. Also, a mass-weighted (sieve) size distribution function is presented for random breakage. Nevertheless, the results refute the random breakage of clinker and other brittle particles after impact loading. Instead, the sieve size distribution of fragments may evolve due to crack branching/merging and Poissonian crack nucleation processes.