Department of Aerospace and
Mechanical Engineering

Direct Numerical Simulations of Multiphase Flows: An Introduction

During the last decade, DNS of multiphase flows has firmly been established as one of the most promising approach to understand the dynamics of complex flows containing two or more co-flowing fluids or phases. For those of us who have been promoting this approach for more than twice that time this is certainly exciting. However, there is still a very high barrier to entry into the field. Numerical methods are becoming increasingly sophisticated and the easy physical problems have been ``taken.'' Increasingly the opportunities for meaningful contributions involve complex physical problems requiring advanced methods and big computers. Thus, in many ways, the path to groundbreaking contributions is now longer than when we first started doing DNS studies of multiphase flows. Although this is a situation that is not unique to our field, I feel that in multiphase flows, which is so central to almost every aspects of the modern way of living, it is particularly urgent to get more researchers looking at more advanced problems. Thus, I have increasingly become concerned with how to bring new researchers quickly to the point where they can start to look at problems that matter.

I have participated in writing two books that are aimed at helping lower the entry barrier. The first one, co-edited with Andrea Prosperetti and containing contributions from several of the world's foremost experts on computational multiphase flows, gives an overview of computational modeling of such flows at different scales. The second book, co-authored with R. Scardovelli and S. Zaleski, describes Volume of Fluid and Front Tracking methods for DNS of multiphase flows and presents results obtained by these methods.

On this page I have gathered additional material that I think may be helpful for someone who is interested in getting into CFD of multiphase flows.

  • A tutorial containing a step-by-step development of a 2D front-tracking code in Matlab. Starting from a simple code for variable density where the advection is done using upwind, to a code using stretched grids, front-tracking for the advection of the material interface and computation of surface tension and a second order ENO scheme for the advection terms.
  • A fairly old 2D front-tracking code written in Fortran 77, along with a few input files.
  • This material is freely available and can be copied and reused for any purpose whatsoever. There is, however, absolutely no guarantee about its correctness or applicability to a specific situation. If you find an error or feel that something could be explained more clearly, then I appreciate hearing from you.

    Some of the material made available here has been developed with support from the National Science Foundation.

    Last updated 2/15/2011 by Gretar Tryggvason