A differentially weighted Monte Carlo method for two-component coagulation (Q2638226)

The paper deals with spatially homogeneous two-component coagulation processes, described by a transport equation for the number of particle pairs having volume in an infinitesimal region of the volume-space, which is an extension of the Smoluchowski equation for one-component coagulation. The simulation is based on a weighted Monte Carlo Method where each simulation particle has assigned a weight representing the number of real particles associated with the considered simulation particle. Two coagulation mechanisms are adopted, \textit{the full coagulation rule} and the \textit{probabilistic coagulation rule}, along with a selection criterium for the coagulation pairs. Moreover, in order to reduce the statistical noise in the less populated region of the volume-space, a shift action is prescribed. These latter allows to keep the number of particles in each cell between a prefixed minimum and maximum value and in the meantime reduces the computational effort. The performance of the simulation scheme is tested in two ideal cases for which an analytical solution is known: an initially monodisperse distribution of each component and an initially polydisperse distribution of each component with a simplified coagulation kernel. As expected, the method proposed by the authors has a better performance with respect to the non weighted Monte Carlo in the regions where only few particles are present, improving the accuracy of the overall distribution properties such as moments.