Application of the multiscale approach to simulation of air sorbent filtration (Q2116725)

The authors consider the problem of numerical modeling air filtration by granular sorbents. The subject of cleaning air is relevant for industrial purification from harmful products of manufacturing, as well as for household cleaning air. A multiscale approach is suggested for solving a compound filtration problem in real geometry conditions. Such approach is based on the method of splitting into physical processes and space-time scales. The multiscale approach allows to compute the evolution of the impurity concentration in regions of varying complexity, including real geometry. From a practical point of view, such numerical algorithms can become a useful tool for the development and optimization of filtration devices, as well as for their efficient exploitation. The basic models on the macroscale are the equations of quasi-hydro- and quasi-gasdynamics and convection-diffusion equations. On the microscale, models of particles, including molecular dynamics, are used. The numerical implementation of the former is based on the finite-volume method on unstructured meshes, and the latter are implemented using the Verlet scheme and its generalizations. An analysis of the results of numerical experiments confirms the efficiency of the proposed multiscale approach.