Many natural and industrial processes involve the flow of solid particles or liquid droplets whose dynamical evolution and morphology are intimately coupled with a carrier gas. The nonlinear and multiscale nature of such flows often precludes a direct analytic solution, and instead we must turn to simulations that rely on subgrid-scale models. The first part of this talk will focus on the flow physics taking place at the microscale (scale of individual particles). A new stochastic drag formulation will be presented that is designed to capture the effect of particle microstructure on velocity statistics. We will then shift focus to modeling such flows at the macroscale (flows made up of millions of particles or more). A new data-driven framework based on sparse regression will be presented for model closure of the multiphase Reynolds Averaged Navier—Stokes (RANS) equations. We will focus on gas-solid flows at moderate volume fractions and Reynolds numbers, such that strong coupling between the phases gives rise to large-scale heterogeneity (clusters) that drive the underlying turbulence. It will be shown that sparse regression can identify compact, algebraic models that respect frame invariance from high-resolution simulation data.

Bio: Jesse Capecelatro is an Associate Professor in the Departments of Mechanical Engineering and Aerospace Engineering at the University of Michigan. Prior to joining Michigan in 2016, he was a postdoc at the University of Illinois Urbana-Champaign, and before that received a Ph.D. from Cornell in 2014. His research is broadly under the realm of fluid mechanics, with an emphasis on multiphase flow, turbulence, reacting flows, and high-performance computing. Applications include renewable energy, propulsion, disease transmission, and space exploration.