Dr. Julian Fox (Dept. of Chemical Engineering, MSU)

11/02/2023  3:10pm

Abstract:

With the advent of supercomputers and parallelization, the field of Computational Fluid Dynamics (CFD) has greater potential than ever to provide profound insights into the underlying physics of fluid dynamics. Using CFD, we can iteratively solve differential equations, oftentimes the Navier-Stokes equations, by discretizing our domain in time and space using numerical methods. As these CFD solvers get more specialized for a given scenario, such as the atomization that occurs in a fuel injector, it is often times required that novel approaches in the form of numerical methods are required to solve the problems in these solvers. Simulations of these gas-liquid flows are often performed using a volume of fluid (VOF) method that uses a Semi-Lagrangian discretization that provides accurate and robust solutions. The current work explores a novel discretization of the Navier-Stokes equations using a Semi-Lagrangian framework which allows for consistency in the discretization of all divergence terms within the Navier-Stokes equations. Additionally, this approach has the potential to lower the computational cost of these VOF simulations for gas-liquid flows.