Proposal Number: 0625138 Principal Investigator: Prosperetti, Andrei Affiliation: Johns Hopkins University Proposal Title: A multi-scale approach to disperse two-phase flow
Intellectual Merit:
This work is the first application of multiscale computing to a disperse multiphase flow problem. Its objectives are to develop methods to couple detailed local numerical solutions of fluid flow with suspended particles to a coarse-grained description based on averaged equations. The specific objective is to study the effect of the bounding solid walls on a disperse fluid particle flow. While much progress has been made in the treatment of such flows by means of averaged equations, the proper boundary conditions to be imposed on the solution of these equations are not well established. In order to address this problem, it is proposed to conduct a multiscale simulation in which the wall region will be treated by direct numerical simulation accurately solving the Navier-Stokes equations with suspended finite-sized particles. This detailed model will be coupled to an averaged-equations model describing the flow in the regions away from the walls. Beyond the specific results for the boundary conditions problem, it is expected that an accurate and efficient solution of the many issues that arise in the coupling of the two descriptions will open the way to other applications of multiscale computing to multiphase flow. Computational limitations have forced the vast majority of previous simulations to be conducted by approximating the particles as mass points in conditions of exceedingly small particle concentration. There are many very important situations which cannot be studied by these means: particles suspended in a liquid, non-dilute systems and many others. The algorithm to be used here is capable of taking full account of the finite extent of the particles and to simulate dense systems.
Broader Impacts:
Multiphase flow is a very vital area of contemporary research. There are compelling scientific reasons for the interest in this research: the statistical treatment of problems of this type is as necessary as it is challenging; a large fraction of the entire discipline of Fluid Mechanics is relevant; and fundamental processes, such as clustering and diffusion are incompletely understood. The successful development of multiscale computational techniques will have a significant impact on the progress of the discipline. Young researchers trained in this field will be able to join a vibrant scientific community addressing significant research objectives.
